among children [1–3], obese children also have an increased risk for coronary
heart disease (CHD) in adulthood. The current chapter reviews the relation of
childhood obesity to various CHD risk factors, atherosclerosis, and clinical disease
in adulthood. Many of the presented data are from the Bogalusa Heart Study
(Louisiana, USA), a long-term study of the early natural history of CHD [4]. This
panel design of this study, which conducted seven examinations of children (ages
5–17 years) and four examinations of young adults (ages 18–37 years) between
1973 and 1995 [5], allows for both cross-sectional and longitudinal analyses.
Obesity and CHD Risk Factors in Childhood
Despite the inherent limitations of weight-height indices in quantifying adipose
tissue, the body mass index (BMI, kg/m2) has been used as a surrogate measure
of obesity among adults for many decades, and is now widely used among
children. However, because BMI levels in early life vary substantially (r 0.5)
with age and height [1], it is necessary to represent BMI levels relative to a
child’s age and sex peers.
BMI levels among children can be classified by estimating levels at age
18 years from current BMI, and then applying the adult cutpoints of 25 kg/m2
(overweight) and 30kg/m 2 (obese) to these extrapolated values [6]. Sex-specific
BMI-for-age centiles and Z-scores have also been developed from US data [7],
and these values will be used throughout this chapter. It has been recommended
that children with a BMI 95th centile of these US data, which roughly corresponds
to an expected BMI between 25 and 30kg/m 2 at age 18 years [6], be
considered overweight. It should be realized, however, that because the relation
of BMI to various diseases differs by ethnicity [8], a single classification scheme
may not be appropriate for all children.
BMI levels among children show a non-linear relation to adverse levels of
lipids, insulin, and blood pressure (fig. 1, upper panel), with the prevalence of
adverse levels increasing markedly at very high BMI levels [9]. In these analyses
of 5- to 17-year-olds (n 23,758) from the Bogalusa Heart Study, risk-factor
cutpoints were chosen so that prevalence of adverse levels of each characteristic
was 5%. BMI was most strongly associated with levels of insulin (r 0.5), and
although the prevalence of adverse levels varied from 1% among the thinnest
children to only 3% at the 85th centile of BMI, almost 40% of the children with
a BMI 99th centile had a high insulin level. Among these severely overweight
children, the prevalence of other risk factors ranged from 15% (low HDL cholesterol)
to 33% (high triglycerides).
Slightly stronger associations were observed with waist girth than for BMI.
For example, whereas levels of HDL cholesterol showed a correlation of r
0.29 with BMI, the correlation with waist girth was r 0.33. Furthermore,
HDL cholesterol levels were more strongly correlated with the subscapular
skinfold thickness (r 0.28) than with the triceps skinfold thickness (r
0.21). Although the importance of visceral fat in these associations is uncertain
[10], it is likely that these contrasting associations reflect some aspect of
body fat distribution.
The relation of childhood obesity to the clustering of multiple risk factors
has also been examined [11], and data from the Bogalusa Heart Study are summarized
in the bottom panel of figure 1. Overall, about 20% of the examined
children had adverse levels of at least one of the five risk factors considered, but
this percentage varied from 10% among the thinnest children to 70% among the
heaviest children. Similar to the findings for the individual risk factors, the associations
were markedly non-linear. The increase was most striking for those with
3 risk factors, with the proportion increasing from 0% (thinnest children), to
1% (85th BMI centile), and to 10% (99th BMI centile). As the presence
of multiple risk factors is strongly associated with the early stages of atherosclerosis
[12], the recent secular increases in the prevalence of overweight
( 95th centile) and severe overweight ( 99th centile) among children [13, 14]
are of particular concern.
Concurrent with these increases in childhood obesity, the prevalence of
type 2 diabetes has also increased substantially among adolescents [15, 16].
Obesity is very common among these newly diagnosed cases, and it has been
found that most of the increase in type 2 diabetes among Pima Indian children
can be accounted for by changes in childhood obesity and intrauterine exposure
to diabetes since 1970 [15].
In addition to these associations, childhood obesity is associated other CHD
risk factors, such as left ventricular hypertrophy, elevated levels of C-reactive
protein, homocysteine and lipoprotein(a), and various pro-coagulant factors
[17, 18]. Furthermore, whereas childhood BMI is inversely associated with levels
of HDL cholesterol and large HDL, it is positively associated with levels of
small HDL [19]. These contrasting associations with HDL subclasses may also
increase the progression of atherosclerosis [20].
Longitudinal Associations with Adult Complications
Risk Factors
It has been estimated that 40% of overweight children will be obese in
adulthood (positive predictive value), while 15 to 20% of obese adults had been
overweight as children (sensitivity) [21, 22]. Although these estimates are
greatly influenced by the cutpoints used to define overweight and obesity [23],
correlational analyses also indicate a moderate degree of tracking for BMI. For
example, over a (mean) 17-year follow-up of 2- to 17-year-olds, the correlation
between childhood and adult levels of BMI was 0.61 [24].
Although adults who had been overweight children have adverse risk factor
levels [25], the persistence of obesity throughout life suggests that these
associations may reflect the importance of adult, rather than childhood, weight
status. This possibility was examined a cohort study of 2609 children (ages 2–17
years) who were followed for an average of 17 years (table 1) [24]. As compared
with adults who had been relatively thin (BMI 50th centile) children, those
who had been overweight in childhood had a 12.4 kg/m2 higher BMI, a 7 mg/dl
lower HDL cholesterol level, and adverse levels of other risk factors (first two
columns).
However, within categories of adults who were normal-weight (adult
BMI 25 kg/m2) or obese ( 30kg/m 2), risk factor levels varied only slightly
according to childhood weight status (final four columns). For example, there
was a 1 mg/dl difference ( 25 kg/m2) or no difference ( 30kg/m 2) in HDL
cholesterol levels according to childhood weight status within categories of
adult BMI. Furthermore, normal-weight adults who had been overweight children
had a lower ( 23 mg/dl, p 0.05) mean triglyceride level than did other
normal-weight adults, likely resulting from a relative decrease in BMI levels
[26] in this group of overweight children who became normal-weight adults.
Overall, these findings indicate that the relation of childhood obesity to adult
risk factors is indirect, resulting from the persistence of childhood obesity into
adulthood.
Atherosclerosis
The initial stages of atherosclerosis are associated with maternal
hypercholesterolemia among neonates [27]. Based on pathology studies of children
and young adults, childhood obesity also appears to be important in the
development of these early lesions. For example, among subjects who died
(ages 2–39 years) from external causes, previously measured BMI was associated
(r 0.24–0.41) with the extent of fatty streaks and fibrous plaques [12].
Furthermore, although some correlations were not statistically significant,
raised lesions in the coronary artery also appear to be associated with subcutaneous
fat (as assessed by the thickness of the panniculus adiposus) and BMI
measured at death among 15- to 34-year-olds [28].
The relation of childhood obesity to atherosclerosis has also been studied
using B-mode ultrasonography, a non-invasive technique that can quantify the
intima-media thickness (IMT) of the carotid artery [29]. Despite various limitations
[30] carotid IMT is thought to be a marker of generalized atherosclerosis,
and IMT among adults is associated with obesity and other CHD risk
factors [31], arteriographically documented coronary artery disease [32], and
subsequent CHD [33]. Furthermore, weight loss decreases the rate of IMT
progression [34].
Although there have been fewer studies of children, hypercholesterolemia
is associated with carotid IMT by age 6 years [35]. Several cross-sectional studies
have also found that obesity in childhood and adolescence is associated with
carotid IMT [36–38], and in most studies the predictive ability of BMI is similar
to that for lipid and lipoprotein levels. Although childhood obesity was not
associated with carotid IMT in some studies [39, 40], it was found to be associated
with other characteristics of the carotid artery, such as wall stiffness or
endothelial dysfunction.
The most interesting use of B-mode ultrasonography, however, is in longitudinal
studies can examine the relation of childhood obesity to carotid IMT
in adulthood. For example, levels of BMI and triceps skinfold thickness among
8- to 18-year-olds were predictive of the adult carotid IMT (ages 33 to 42 years)
in the Muscatine Heart Study [41]. Although the longitudinal associations with
carotid IMT were statistically significant only among women (r 0.18 for
childhood BMI), these relationships over an approximately 25-year follow-up
period are noteworthy. These findings also extend an earlier report from the
same group [42] showing that childhood weight is related to coronary artery
calcification (assessed by electron beam computed tomography) in adulthood.
In this earlier study, the observed associations between childhood weight and
the presence of adult calcification were slightly stronger among men (odds
ratio 2.9) than among women (odds ratio 2.1).
It would be very interesting to determine if the longitudinal relations of
childhood obesity to carotid IMT and coronary artery calcification in adulthood
were mediated by the persistence of childhood obesity into adulthood, or if the
effects of childhood obesity were independent of adult weight status.
0 komentar:
Posting Komentar