Cholesterol
Screening
in Pediatric
Practice
David C. Goff, Jr, MD*; Ge A. Donker, MD, MPH*; Jesse D. Ragan, Jr,
MDt; A. Thomas Adkins, MDt; R. Peter Killinger, MDf;
John W. Caudill, Jr, MDf; and Darwin R. Labarthe, MD, PhD*
From the *Epidemiology Research Center, The University of Texas Health Science Center at Houston, School of Public Health, and rPrivate Pediatric Practice, Houston
ABSTRACT. Four pediatricians introduced a portable
cholesterol analyzer into their group practice. Their ex-perience is described on the basis of 12 months of screen-ing in 1665 children and adolescents. The overall 50th and 90th percentile values for a subgroup of 1406
rou-tinely screened children were 156 and 197 mg/dL,
re-spectively, but there was marked variation in these values among specific age and sex groups. Cholesterol levels decreased by age group during the early teenage years and increased thereafter, these changes occurring at ages approximately 2 years younger for girls than for boys. Further analysis of screening results for 398 sibling pairs demonstrated significant concordance between paired cholesterol levels when classified by the respective age-and sex-specific 90th percentile values for each member of the pair. Sibling pairs in which both members’ choles-terol values exceeded their 90th percentile value were identified 2.4 times as frequently as expected (confidence interval 1.1 to 4.5, P = .029). The observations reported
here indicate that office-based cholesterol screening in a pediatric practice may be both practical and useful, al-though further consideration of screening criteria is needed. Age- and sex-specific reference values for choles-terol levels during childhood could improve screening results. Special emphasis should be directed toward screening siblings of children in whom high cholesterol levels have been detected. Pediatrics 1991;88:250-258;
cholesterol screening, children, adolescents.
ABBREVIATIONS. LDL, low-density lipoprotein; HDL, high-density lipoprotein.
Hypencholesterolemia is an important factor in the development of atherosclerosis. Screening of
Received for publication May 29, 1990; accepted Jul 23, 1990. Reprint requests to (D.R.L.) Southwest Center for Prevention Research, Reuel A. Stallones Bldg, 1001, 1200 Herman Pressler Dr, Houston, TX 77030.
PEDIATRICS (ISSN 0031 4005). Copyright © 1991 by the American Academy of Pediatrics.
adults for hypercholesterolemia has become ac-cepted in populations with high coronary disease mortality; however, screening of children has been
more controversial.
Autopsy studies have confirmed the presence of fatty streaks and raised lesions in the aortas and coronary arteries of teenagers and young adults.13 Antemortem total and low-density lipopnotein
(LDL) cholesterol levels among decedent teenagers in the Bogalusa Heart Study were positively come-lated with the degree of involvement of the aortic
wall with fatty streaks. The group with cholesterol levels between 140 and 170 mg/dL had
approxi-mately 25% involvement, whereas the group with levels greater than 200 mg/dL had approximately 50% involvement.4 Furthermore, total cholesterol levels have been shown to track within childhood and adolescence, and from childhood to adult-hood.5’#{176} The detection of high cholesterol levels in
children might therefore allow early initiation of interventions in a group at especially high risk for later complications of atherosclerosis.
The American Academy of Pediatrics has me-cently expressed the view that the disadvantages and limitations of screening of the general popula-tion of children outweigh the advantages.” The limitations cited by the Academy included poor
standardization of equipment, diurnal and seasonal variations in cholesterol levels, and the possibility of inappropriate use of drug therapy on the basis of a single sporadically elevated cholesterol level. Ad-ditionally, the Academy raised concerns that die-tary or drug therapy for hypercholesterolemia
cholesterol screening, but only for those children, older than 2 years of age, who have a positive family
history of dyslipidemia or early myocardial infarc-tion in a first-degree relative.
The American Health Foundation has also
pub-lished screening recommendations recently, in an extensive monograph addressing this and related issues.’4 Their recommendations included changes in dietary habits aimed at reducing the US
popu-lation mean of cholesterol levels for children from the current 160 to 140 mg/dL. This target choles-terol level is described as being wholly compatible with normal growth and development, on the basis of observations in childhood populations. If only
those children with a positive family history of dyslipidemia or early myocardia! infarction were screened, a recent report shows, a large proportion of all children with elevated cholesterol levels would remain undetected.’5 On the basis of such consid-erations, the American Health Foundation
recom-mended screening of all children older than 2 years of age for total cholesterol, as an integral part of a broader cardiovascular disease risk assessment and modification strategy.’4
With the advent of new laboratory technology and conflicting recommendations in the literature, interest in the results of office-based cholesterol
screening is growing. Four private pediatricians in the Gallenia area of Houston (coauthors J.D.R.,
A.T.A., R.P.K., and J.W.C.) have been performing capillary blood cholesterol determinations during well-child visits since March 1988. The experience
of 12 months of screening, allowing for a 3-month run-in period, is reported here, including an inves-tigation of the patterns of cholesterol levels across
the age range from 4 through i9 years and an examination of the concordance of high cholesterol
levels between siblings.
METHODS
Study Design
This report is based on retrospective review of pediatric office records for the purpose of describing
the experience of an outpatient pediatric practice, with respect to screening cholesterol levels. The four pediatricians receive approximately 28 600
of-fice visits per year. The office personnel include four nurses, three receptionists, and two
book-keepers. The practice population is predominantly white and middle and upper class. The group began cholesterol screening in March i988, after a
repre-sentative of the instrument manufacturer in-structed the nurses in procedures for use and cali-bration of the Reflotron (Boehninger Mannheim
Diagnostics) device for biochemical analysis. The
nurses performed all office laboratory procedures. One physician saw approximately 14% of the
pa-tients and only infrequently performed cholesterol determinations. A second physician saw 23% of the patients and performed cholesterol determinations only when requested to do so by a parent. The two remaining physicians saw 27% and 36% of the patients, respectively. They both performed choles-terol determinations routinely, at well-child visits, for patients 4 years old or older. The more detailed aspects of the following analyses are based on
ob-servations among the patients of these last two physicians, after exclusion of data for the highly selective groups screened by the other two
physi-cians.
Subjects
Patients aged 4 through i9 years who had capil-lary blood cholesterol determinations during well-child visits between June i, 1988 and May 31, 1989, were identified from entries in a procedure log
maintained by office staff for each occasion of use of the Reflotron. The patients’ names and exami-nation dates were abstracted from this log. The resulting roster was then alphabetized, and the corresponding office records were reviewed. Sib-lings were linked through coded identification num-bers assigned at the time of this chart review. Sibling identification was verified by checking in-formation concerning parents’ names and addresses recorded in the charts.
Data Collection and Analysis
Data collected from office records included date of birth, gender, race and, for each examination within these 12 months, the date of examination,
screening cholesterol level, systolic and diastolic blood pressure, height, weight, physician, reason for visit, and (if subsequently obtained) lipid profile.
Cholesterol levels were determined on samples of capillary blood that were obtained by finger stick and processed on a Reflotron device whose calibra-tion was performed approximately twice monthly. Causal blood pressure was determined by the phy-sician, with either an aneroid or a mercury sphyg-momanometer. Height and weight were measured
in light clothing, without shoes, with an Accustat stadiometer and a Healthmeter beam balance scale.
The reason for the visit was recorded, and this information was abstracted for the present study to ensure that only well-child visits were analyzed.
Laboratory, a commercial laboratory that uses a Hitachi 736 analyzer to perform total cholesterol (TC) and triglyceride (TG) determinations. High-density lipoprotein (HDL) is measured after dex-tran-magnesium precipitation of other serum
lipo-proteins. Low-density lipoprotein (LDL) is calcu-lated by the equation: LDL = TC - (HDL + TG/
5). MPC Laboratory is certified in cholesterol
meas-urement through the use of certified reference ma-tenials provided by the College of American Pa-thologists.
Data collection and editing were performed with a Toshiba “lap-top” computer for on-site data entry and DBase III software. Data analysis was accom-plished with SPSS-PC. The primary analyses in-cluded the following:
1. An analysis of variance procedure was per-formed to investigate the effect of physician selec-tion practices on the average cholesterol value found in a population of children and adolescents. This analysis examined the entire population of 1665 patients screened by the four physicians.
2. The cholesterol patterns in males and females from ages 4 through 19 years were next described. For this analysis, only the 1406 patients of the two physicians who used routine screening were exam-med. The total group was first stratified by sex. Then each child was assigned to 2-year overlapping age groups (that is, a 5-year-old would be classified both as 4 through 5 and 5 through 6 years old; children aged 4 or 19 years were included only once). This procedure improved the stability of the descriptive statistics; it also smoothed the curves of cholesterol by age (see “Discussion”).
3. The degree of sibling concordance of elevated
cholesterol levels was then analyzed in 398 sibling pairs. For those families with at least two offspring screened during the year, the concordance between
siblings with respect to elevated cholesterol levels was determined. In this analysis, both the Lipid
Research Clinics’ overall 90th percentile value for children aged 0 through 19 years and the age- and sex-specific 90th percentile value calculated from the present study population were used as criteria
of an elevated cholesterol concentration.’6 Two-by-two tables were constructed, and a
x2
statistic was computed for each table. The number of sibling pairs in which both siblings exceeded their nespec-tive age- and sex-specific 90th percentile cholesterol values was divided by an expected number calcu-lated under the assumption of total independencebetween siblings. The significance of this preva-lence ratio was determined assuming a Poisson probability distribution. A one-sided test of signif-icance and a corresponding 90% confidence interval were calculated.
4. Finally, the follow-up of 104 patients with elevated cholesterol levels was described. For this analysis, cholesterol measurements that exceeded the Lipid Research Clinics’ sex-specific 90th
per-centile values for children aged 0 through 19 years were considered elevated.’6 The individual 6-month follow-up experience of each of the 104 patients found to have elevated cholesterol measurements within the first 6 months of the observation period
was described.
Further analyses, which address relationships be-tween cholesterol levels and anthropometnic
van-ables and the follow-up of patients with elevated cholesterol levels, are in progress and will be
me-ported separately.
RESULTS
Study Group Selection
A total of 1665 children and adolescents (936 boys and 729 girls), in the age group 4 through 19
years, had an initial cholesterol measurement per-formed during well-child visits between June 1, 1988, and May 31, 1989. Table 1 shows the age and sex distributions of patients screened per physician and overall. Table 2 shows the means and standard
deviations of age and cholesterol concentration for patients screened by each physician and for the
entire group. Analysis of variance of cholesterol values by physician, adjusted for age, showed a significant difference among the values of choles-terol concentration by physician (P = .015). This
difference could be attributed entirely to physician 2, whose patients had a higher mean value of cho-lesterol (163 mg/dL) than the two (physicians 3
and 4) who ordered cholesterol screening routinely (mean values 157 and 158 mg/dL, respectively). Physician 1 screened only 20 patients, too few to
justify further analysis. Inasmuch as there was no difference in mean cholesterol values between phy-sicians 3 and 4, the remainder of the analysis was restricted to those 1406 subjects (791 boys and 615 girls), whose mean ages were 10.9 and 10.8 years, and whose mean cholesterol values were 155.6 and 159.0 mg/dL, respectively.
Figure 1 shows the sex-specific cholesterol distni-butions for this study group of 1406 patients. (The
cholesterol levels represented at i05 mg/dL include those for all patients with cholesterol values below the detection limit of the measurement device, or
popu-TABLE 2. Mean Values (and Standard Deviations) of Age and Cholesterol Concentration, by Physician and Overall
Physician No. of Children
Age, y Cholesterol, mg/dL
1 20 14.1 (2.3) 148 (28.2)
2 239 11.7* (3.0) 163t (34.5)
3 671 11.4 (3.9) 157 (30.7)
4 Total
735 1665
10.4 (3.9) 11.0 (3.8)
158 (31.1)
158 (31.4)
* < .001 when compared with mean age of patients of physicians 3 and 4.
t P = .007 when compared with mean cholesterol value
of patients of physicians 3 and 4.
16
Boys
P 14
-
- GirlsE
R
C 12 E
N T 10 0
F
8-C
H
6-jJ
_TI1IjF__
105’ 125 145 165 185 205 225 245 265 285
CHOLESTEROL LEVEL (mg/dl)
Fig 1. Distribution of cholesterol values for boys and
girls. *The interval centered at 105 mg/dL includes all
determinations below the detection limit ofthe Reflotron,
or 100 mg/dL.
TABLE 1. Childre n 5cr eened per Physician, and 0 verall, by Age and Sex
A ge, y Physician
Total
1 2 3 4
M F M F M F M F M F Total
4 1 2 2 3 2 5
5 2 3 28 20 40 36 70 59 129
6 3 3 16 25 46 37 65 65 130
7 4 7 31 14 41 34 76 55 131
8 5 9 30 23 27 20 62 52 114
9 16 5 19 22 37 25 72 52 124
10 1 15 11 32 30 33 28 81 69 150
11 13 9 32 15 29 23 74 47 121
12 2 3 24 18 26 22 25 21 77 64 141
13 2 14 13 40 29 23 23 79 65 144
14 2 2 15 8 34 22 31 20 82 52 134
15 1 2 6 8 28 17 20 20 55 47 102
16 2 11 4 24 11 23 12 60 27 87
17 1 1 6 20 16 13 12 35 34 69
18 1 4 2 16 22 15 11 35 36 71
19 1 4 2 4 1 9 3 12
20 1 1 1
Total 12 8 133 106 381 290 410 325 936 729 1665
lation. The mean cholesterol levels for boys (i55.6)
and girls (159.0) were significantly different when analyzed by analysis of variance (P = 0.045). It was
also known from previous studies that boys and
girls in the age group under investigation differ in
their age trends of cholesterol levels.’7’18 For these reasons, further analyses were stratified by both
sex and age.
Of these i406 patients, only 43 (3.1%) were not
white. Analysis of variance of cholesterol level by
ethnicity, adjusted for age and sex, did not show a significant difference in mean cholesterol level
be-tween white patients and others (158.0 vs 168.1 mg/
dL, P = .068). Because nonwhite patients were few
in number, no further analysis of the effect of
ethnicity was attempted.
Changes of Cholesterol With Age
Table 3 and Figures 2 and 3 show the 10th, 50th,
and 90th percentile values of cholesterol for gender-specific 2-year overlapping age groups. Inspection
of the figures reveals that the cross-sectional nela-tionship between age and cholesterol levels was
nonlinear. For both boys and girls, cholesterol levels
tended to be steady during the prepubertal years,
to drop during puberty, and then to rise again during adolescence. Cholesterol levels started to
decrease by age group earlier in girls (9 through 10
years of age) than boys (ii through 12 years of age); they started to increase again earlier as well,
CHOLESTEROL (mg/dl)
250
-.- lot,, Percentile
- 90th Percentile
-
50th Percentile-
LRC 90th Percentile 210k185-
--
170-13O
,
- -- .TABLE 3. Tenth, 50th, and 90t
Overlapping Age Groups*
h Percentile Chol esterol Levels for Sex-Specific 2-Year
Age No. in Age-Group, y Sex Group
10th Percent-ile, mg/dL 50th Percent-ile, mg/dL 90th Percent-ile, mg/dL
M F M F M F M F
4-5 71 58
5-6 130 118
6-7 134 110
7-8 129 91
8-9 113 90
9-10 121 105
10-11 126 96
11-12 112 81
12-13 114 95
13-14 128 94
14-15 113 79
15-16 95 60
16-17 80 51
17-18 64 61
18+ 40 36
122 119 122 116 125 118 125 118 126 122 126 133 125 130 124 111 122 117 117 116 106 112 107 114 113 117 110 114 111 112 161 158 160 160 158 161 158 160 161 161 160 167 156 165 162 163 158 153 150 147 142 147 141 146 140 153 138 155 141 157 208 210 200 204 193 201 201 202 203 206 203 211 199 200 201 197 203 204 193 194 188 193 188 176 181 180 174 192 183 213
* Data from the patients of physicians 3 and 4 only.
90 - -
-
- ----
.-
--4-5 6-7 8-9 10-11 12-13 14-15 16-17 18-19
AGE GROUP
Fig 2. Tenth, 50th, 90th, and Lipid Research Clinics
(LRC) 90th percentile cholesterol levels for boys, for
2-year overlapping age groups.
50th percentile values of cholesterol for each of the age groups tended to be higher for girls.
Cholesterol Concordance Between Siblings
The results of cholesterol screening were further examined by evaluating 398 sibling pairs. Of the 61 families having more than 2 offspring in the study, the 2 oldest were selected for this analysis. When these sibling pairs were classified using the Lipid Research Clinics’ 90th percentile value, i85 mg/ dL,’6 as the criterion for all ages and both sex groups, there was no significant concordance be-tween the siblings
(x2,
1 df = 0.58, P > .25). How-ever, when the sibling pains were reclassified ac-cording to the age- and sex-specific 90th percentile values observed in the group of 1406 patients,sig-nificant concordance between siblings was found
(x2,
1 df = 6.92, P .009). Siblings of patientswhose cholesterol levels exceeded their sex- and age-specific 90th percentile values were found to have corresponding elevations of cholesterol 2.4 times as frequently as would be expected if the cholesterol values of siblings were independent (P
= 0.29, 90% confidence interval = 1.1 to 4.5; see
also Table 4). This result both underscores the importance of recognizing variation in cholesterol values by age and sex and lends support to the
screening of siblings, once a relatively high choles-terol value has been found in a given child.
Follow-up of Patients With High Cholesterol
Values
Within the 12-month period of this study, 138 of 791 boys (17.5%, mean age = 9.8 years) had
choles-terol screening results that exceeded the 90th per-centile value for boys of 185 mg/dL based on data from the Lipid Research Clinics; for 100 of 615 girls (16.3%, mean age = 9.9 years), the screening results
exceeded the corresponding 90th percentile value
for girls of 190 mg/dL.’6 The mean ages of these two groups of patients were lower than those of the larger subgroups of the i406 patients who under-went routine screening (mean ages of boys and girls 10.9 and 10.8 years, respectively; P = .001 and P =
.038). Further analysis was done on the subset of these 238 patients who were identified in the first
6 months of the study, for whom there had been sufficient time for possible follow-up. Of 108 such
CHOLESTEROL (mg/dt)
250-
--- - --
---
-
________
21 0
10th Percentile 50th Percentile 90th Percentile LRC 90th Percentile
19O
---.
- --
-ii.,
170
---130 - - .
9o---
----.---
--- - . ---. .4-5 6-7 8-9 10-11 12-13 14-15 16-17 18-19
AGE GROUP
Fig 3. Tenth, 50th, 90th, and Lipid Research Clinics (LRC) 90th percentile cholesterol levels for girls, for 2-year overlapping age groups.
66 such girls, 25 (37.9%) received follow-up by a second cholesterol check-up, a lipoprotein analysis, on both. In the overall subgroups whose cholesterol values exceeded the sex-specific 90th percentile
val-ues, 37.4% were followed up within 6 months, and 43.4% of those exceeding the 95th percentile were followed up within this same time period.
Lipoprotein analysis was performed for 65 of
these patients during the study period. The sex distribution (35 boys and 30 girls) was similar to
that of the total population, but the children se-lected for lipoprotein analysis were younger than the remainder. Mean values (and standard devia-tions) for LDL and HDL cholesterol were 135.9 (21.1) and 55.3 (12.7) mg/dL, respectively.
Low-density lipoprotein cholesterol concentrations ranged from 93 to 186 mg/dL; HDL cholesterol
concentrations ranged from 32 to 92 mg/dL. The LDL values exceeded the Lipid Research Clinics’ 90th percentile value of 130 mg/dL in 39 (60.0%) of these 65 patients, whereas only 3 patients (4.6%)
had an HDL value below the standard 10th per-centile of 35 mg/dL.’6 In 2 patients (3.1%) the
elevated cholesterol level was due exclusively to an elevated HDL (above the 95th percentile). Of these 65 patients, 16 (24.6%) had LDL/HDL ratios greater than 3.0, and 8 (12.3%) had ratios greater than 4.0.
DISCUSSION
Cholesterol screening was implemented in this office practice with very little disruption and with-out addition of personnel. Since January 1, 1990,
this practice has been participating in a monitoring program for physician office laboratories. This pro-gram requires documentation of daily calibration of
equipment and periodic measurement of unknown samples prepared by the College of American
Pa-thologists. These additional requirements have not led this practice to discontinue cholesterol screen-ing or to hire additional personnel. Screening was applied routinely by two of the four pediatricians
. and accepted by their parents.
--- Several factors are important for the evaluation of a screening procedure. The screening process
should be simple, inexpensive, relatively pain-free (noninvasive), reliable, and accurate. In addition, screening must be based on sound scientific knowl-edge of both the underlying disease process and the appropriate response to the results of the test. In the case of capillary cholesterol measurements, a
simple finger-stick sample of blood, collected in a capillary tube, is the required sample. This method
of sample collection has long been accepted for neonatal screening for metabolic disorders and for monitoring blood glucose levels in patients with
diabetes mellitus.
Concerns have been expressed about the reliabil-ity and validity of capillary cholesterol measure-ments, most recently by the American Academy of Pediatrics.” Laboratory performance standards for
cholesterol measurements have been established by the Laboratory Standardization Panel of the Na-tional Cholesterol Education Program.’3 Portable cholesterol analyzers, such as the Reflotron, have been shown to be capable of meeting these
stand-ards in field trials,’#{176}25 These results are in accord with our own detailed evaluation of the Reflotron
device. Systematic evaluation of the reliability and validity of cholesterol screening in relation to si-multaneous and follow-up determinations in this practice setting is being planned on the basis of this initial assessment.
Screening is indicated only when responses to the test result are planned. In this practice, the physicians reported the intention to confirm ele-vated levels, greater than 200 mg/dL, and to give dietary advice based on literature obtained from the American Heart Association. The decision on the part of two of these physicians to screen children
routinely reflects their judgment that cholesterol screening coupled with dietary advice can be effec-tive, as proposed by the American Health Founda-tion.’4 More widespread application of cholesterol
screening in children may await further evidence of efficacy.
The age patterns of cholesterol levels (as seen in
TABLE 4. Prevalence Ratios for Sibling Pairs Classified by Age- and Sex-Specific 90th
Percentile Cholesterol Levels
Sibling Pairs Observed Expected* Prevalence Ratio
Both >90th percentile 7 2.9 2.4t
One >90th percentile 54 62.2 0.9
Both 90th percentile 337 332.9 1.0
* The expected values were determined assuming total independence of cholesterol levels
within sibling pairs.
t P = .029.
differences from each i-year age group to the next. (Even the single year-of-age data based on
cross-sectional observations fail to reveal the marked heterogeneity among individuals in the actual lon-gitudinal change in cholesterol levels.)
The use of a fixed criterion value for cholesterol screening for each sex group across the preadult age range is currently proposed by the Lipid Research Clinics and the American Academy of Pediat-rics.”6 This practice leads to relative undeniden-tification of high-risk children during puberty, when naturally occurring cholesterol levels are
lower than in the preceding years. In this respect, the use of age- and sex-specific standard percentile
grids could improve the results of cholesterol screening during childhood. Pediatricians are fa-miliar with standard grids to plot head
circumfer-ence, height, and weight of children. These grids have traditionally been used as indicators to assess problems in childhood.
A grid for serum cholesterol level could become a part of each child’s permanent health record in addition to height and weight percentile grids. A plotting of risk factor levels oven time provides an
effective visual assessment of a child’s risk factor profile. The use of such a standard percentile grid for serum total cholesterol level in childhood has
been recommended by other authors as well.14’26 The data most appropriate for constructing such reference values, like those for physical growth, are longitudinal in nature and require frequent, re-peated measurements over many years. Such data should be collected in appropriately designed lon-gitudinal studies.
Familial clustering of hypencholesterolemia has been described in several studies.2738 The present study confirms the findings of familial clustering
described in previous studies and suggests a prac-tical implication for routine screening. A concord-ance between sibling pairs was found, when mdi-viduals were assigned according to sex- and age-specific 90th percentile values derived from this practice setting. The relative risk for elevated cho-lesterol levels of 2.4 for siblings of children with
elevated cholesterol levels in consistent with a
mel-ative risk of 2.7 from the Lipid Research Clinics
Family Study and other studies.31’32 Screening pro-grams could appropriately emphasize testing the
siblings of children with high cholesterol levels, inasmuch as the yield of such testing is high. The concept of the family as the unit of intervention
also supports this approach.
Both genetic factors and common environment contribute in varying degrees to the phenomenon of familial aggregation.33’34 Socioeconomic factors
have been described as influencing familial aggre-gation of serum total cholesterol levels in several
studies. In the East Baltimore Study, a greater sibling concordance was reported in groups with higher socioeconomic status, regardless of race;
however, the Evans County Study reported greater familial aggregation in a lower socioeconomic
group.35’36 The previously described high socioeco-nomic status of this population may have influ-enced the sibling concordance found in the present
study; however, there is insufficient evidence in existing literature to justify distinguishing between
socioeconomic groups when considering the impli-cations of familial aggregation for cholesterol
screening in a population.
Furthermore, the results showed a higher mean value of cholesterol among children screened only
on their parents’ request. The difference shown in Table 2 underestimates the actual effects of selec-tive screening, inasmuch as the patients of
physi-cian 2 were older (mean age = 11.7 years) than the
patients of physicians 3 and 4 (mean age = 10.9
years, P < .001). Screening of children on their parents’ request may well identify those with a positive family history for cardiovascular disease or
dyslipidemia. Those children may, therefore, be at
a higher risk for hypercholesterolemia, but they represent only a small proportion of those found by
nonselective screening.
These data do not address the efficacy of
inter-ventions to lower serum cholesterol concentrations in childhood, nor do these data represent either a cost-effectiveness analysis or a validity assessment
do, however, address several issues of importance with respect to cholesterol screening in childhood:
Routine screening of cholesterol levels can be
implemented in a pediatric practice by using port-able cholesterol analyzers.
The natural changes that occur in cholesterol levels during growth and development should be appreciated when classifying children for further follow-up on evaluation. Age grids for cholesterol percentiles could be developed to allow pediatni-cians to detect high cholesterol levels in children. New longitudinal studies will be required to collect the data most desirable for constructing such grids. Special emphasis should be directed toward screening siblings of children with elevated choles-terol levels.
ACKNOWLEDGMENTS
This work was supported by the Southwest Center for
Prevention Research through grant R48/CCR 60217-02 from the Centers for Disease Control.
We thank the pediatricians’ office staff for their
as-sistance in locating records.
REFERENCES
1. Enos WF, Holmes RH, Beyer J. Coronary disease among United States soldiers killed in action in Korea: preliminary
report.JAMA. 1953;152:1090-1093
2. McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coro-nary artery disease in combat casualities in Vietnam. JAMA.
1971;216:1185-1187
3. Strong JP, McGill HC. The pediatric aspects of atheroscle-rosis. J Atheroscier Res. 1969;9:251-265
4. Newman WP, Freedman DS, Voors AW, et a!. Relationship of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J
Med. 1986;314:138-144
5. Clarke WR, Schrott HG, Leaverton PE, Connor WE, Lauer RM. Tracking of blood lipids and blood pressure in school age children: the Muscatine Study. Circukition.
1978;58:626-634
6. Lauer RM, Lee J, Clarke WR. Factors affecting the rela-tionship between childhood and adult cholesterol levels: the Muscatine Study. Pediatrics. 1988;82:309-318
7. Webber LS, Cresanta JL, Voors AW, Berenson GS. Track-ing of cardiovascular disease risk factor variables in school-age children. J Chronic Di.s. 1983;36:647-660
8. Freedman DS, Shear CL, Srinivasan SR, Webber LS, Ber-enson GS. Tracking of serum lipids and lipoproteins in children over an 8-year period.: the Bogalusa Heart Study.
Prey Med. 1985;14:203-216
9. Webber LS, Cresanta JL, Croft JB, Srinivasan SR, Beren-son GS. Transitions of cardiovascular risk from adolescence to young adulthood-the Bogalusa Heart Study, II: altera-tions in anthropometric blood pressure and serum lipopro-tein variables. J Chronic Dis. 1986;39:91-103
10. Webber LS, Srinivasan SR, Berenson GS. Tracking of serum lipids and lipoproteins over 12 years into young adulthood: the Bogalusa Heart Study. Circulation. 1988;78(suppl 2):481. Abstract
11. Committee on Nutrition, American Academy of Pediatrics. Indications for cholesterol testing in children. Pediatrics.
1989;83:141-142
12. The Expert Panel. Report of the National Cholesterol Ed-ucation Program Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults. Arch Intern
Med. 1988;148:36-69
13. Current status of blood cholesterol measurement in clinical laboratories in the United Stats: A report from the Labora-tory Standardization Panel of the National Cholesterol Ed-ucation Program. Clin Chem. 1988;34:193-201
14. Wynder EA, ed. Coronary artery disease prevention: choles-terol, a pediatric perspective. Prey Med. 1989;18:323-409 15. Garcia RE, Moodie DS. Routine cholesterol surveillance in
childhood. Pediatrics. 1989;84:751-755
16. Rifkind BM, Segal P. Lipid Research Clinics Program
ref-erence values for hyperlipidemia and hypolipidemia. JAMA. 1983;250:1869-1872
17. Berenson GS, Srinivasan SR, Cresanta JL, Foster TA, Web-ber LS. Dynamic changes of serum lipoproteins in children during adolescence and sexual maturation. Am J Epidemiol. 1981;113:157-170
18. Lauer RM, Lee J, Clarke WR. Factors affecting the rela-tionship between childhood and adult cholesterol levels.
Pediatrics. 1988;82:309-318
19. Sedor FA, Holleman CM, Heyden S, Schneider KA.
Reflo-tron cholesterol measurement evaluated as a screening tech-nique. Clin Chem. 1988;34:2542-2545
20. Boerma GJM, Gelderland J, van Gorp I, Leijnse B. Use of the Reflotron system for cholesterol assay in capillary blood, venous blood, and serum: evaluation of accuracy and lot-to-lot reagent comparability. Clin Chem. 1988;34:
2117-2122
21. James R, Tyler C, Henrikson D. An evaluation of the accuracy of the Reflotron system in the field. Med J Aust. 1988;149:130-131
22. Burke JJ, Fischer PM. A clinician’s guide to the office measurement of cholesterol. JAMA. 1988;259:3444-3448
23. Lefebvre RC, Lasater TH, McKinlay SM, Gans KM, Walker
N, Carleton RA. Performance characteristics of a blood cholesterol measuring instrument used in screening pro-grams. Public Health Rep. 1989;104:266-270
24. Bachorik PS, Bradford RH, Cole T, et al. Accuracy and precision of analyses for total cholesterol as measured with the Reflotron cholesterol method. Clin Chem.
1989;35:1734-1739
25. Greenland P, Bowley NL, French CA, Meiklejohn B, Gag-liano S, Sparks CE. Precision and accuracy of a portable blood analyzer system during cholesterol screening. Am J Public Health. 1990;80:181-184
26. Cresanta JL, Hyg MS, Burke GL, Downey AM, Freedman DS, Berenson GS. Prevention of atherosclerosis in child-hood. Pediatr Clin North Am. 1986;33:835-858
27. Johnson BC, Epstein FH, Kjelsberg MO. Distribution and familial studies of blood pressure and serum cholesterol levels in a total community: Tecumseh, Michigan. J Chronic
Dis. 1965;18:147-160
28. Hennekens CH, Jesse MJ, Klein BE, Gourley JE, Blumen-thal S. Cholesterol among children of men with myocardial infarction. Pediatrics. 1976;58:211-217
29. Lee J, Lauer RM, Clarke WR. Lipoproteins in the progeny of young men with coronary artery disease: children with increased risk. Pediatrics. 1986;78:330-337
30. Schrott HG, Clarke WR, Wiebe DA, Connor WE, Lauer RM. Increased coronary mortality in relatives of hypercho-lesterolemic school children: the Muscatine Study.
Circula-tion. 1979;59:320-326
31. Morrison JA, Namboodiri K, Green P, Martin J, Glueck CJ. Familial aggregation of lipids and lipoproteins and early identification of lipoproteinemia: the Collaborative Research Clinics Family Study. JAMA. 1983;250:
1860-1868
32. Morrison JA, Kelly KA, Mellies MJ, Groot I de, Glueck CJ. Parent-child associations at upper and lower ranges of plasma cholesterol and triglyceride levels. Pediatrics. 1978;62:468-477
environmental influence on cardiovascular risk factors for children. Am J Epidemiol. 1982;116:385-393
34, Sing CF, Orr JD. Analysis of genetic and environmental sources of variation in serum cholesterol in Tecumseh, Michigan, IV: separation of polygene from common environ-mental effects. Am J Hum Genet. 1978;30:491-504
35. Van Natta P, Jenkins J, Chase G, Alvaro K, Bachorik PS, Kwiterovich P0 Jr. The East Baltimore Study, II: familial
aggregation of plasma lipids in juveniles in a black inner city population. Am J Epidemiol. 1981;114:385-397
36. Tyroler HA, Heiss G, Heyden 5, Hames CG. Family follow-up study of serum cholesterol in Evans County, Georgia. J
Chronic Dis. 1980;33:323-330
37. Chase GA, Kwiterovich P0, Bachorik PS. The Columbia Population Study, II: familial aggregation of plasma choles-terol and triglycerides. Johns Hopkins Med J. 1979;145:150-156
38. Shear CL, Frerichs RR, Weinberg R, Berenson GS. Child-hood sibling aggregation of coronary artery disease risk factor variables in a biracial community. Am J Epidemiol. 1978;107:522-528
BABES IN ADLAND
One simple gimmick helped “Look Who’s Talking” become a movie
block-buster last year: The flick’s star, a baby, “talked” throughout the movie in the voice of an adult actor, Bruce Willis. Only the audience, the twist went, could
hear the baby speaking.
Madison Avenue thought the gag was hot, too: The “baby-talk” device is now showing up in ads for everything from Kellogg’s Eggo waffles to Burger King sandwiches. Baby conversations have also made their way into ads for Scotts’ Baby Fresh baby wipes, Fuji Photo film and even a maternity ward at a hospital
near Boston.
. . . the idea of using babies in ads isn’t new. Advertising executives, armed
with the hunch that babies have a strong emotional pull on consumers, have
cast babies in starring roles in commercials for years, including those for Michelin tires, Du Pont Stainmaster carpet and Angel Soft toilet tissue. But the popularity of “Look Who’s Talking,” advertising executives say, has put a new spin on the old baby-in-ad formula, letting the baby speak for himself.
1991;88;250
Pediatrics
Adkins, R. Peter Killinger and John W. Caudill, Jr
David C. Goff, Jr, Ge A. Donker, Darwin R. Labarthe, Jesse D. Ragan, Jr, A. Thomas
Cholesterol Screening in Pediatric Practice
Services
Updated Information &
http://pediatrics.aappublications.org/content/88/2/250
including high resolution figures, can be found at:
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml
entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or in its
Reprints
http://www.aappublications.org/site/misc/reprints.xhtml
1991;88;250
Pediatrics
Adkins, R. Peter Killinger and John W. Caudill, Jr
David C. Goff, Jr, Ge A. Donker, Darwin R. Labarthe, Jesse D. Ragan, Jr, A. Thomas
Cholesterol Screening in Pediatric Practice
http://pediatrics.aappublications.org/content/88/2/250
the World Wide Web at:
The online version of this article, along with updated information and services, is located on
American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.
