Challenges
to Implementing
the
Current
Pediatric
Cholesterol
Screening
Guidelines
Into
Practice
Barbara A. Dennison, MD; Paul L. Jenkins, PhD; and Thomas A. Pearson, MD, PhD
ABSTRACT. Objective. The Expert Panel on Blood
Cholesterol Levels in Children and Adolescents of
the National Cholesterol Education Program (NCEP) rec-ommends selective screening of children for high blood cholesterol. We determined the number of children, who, according to the guidelines, should be targeted for cholesterol screening.
Design. Population survey.
Setting. Permanent household residents in Otsego
County, NY.
Participants. Total population-based sample of
17 444 households (86.6% response rate) including 44 565
participants, of whom 10 457 were children, aged 2
through 19 years.
Main outcome measures. Percent of children
qualify-ing for cholesterol screening under the NCEP Children’s Panel guidelines.
Results. Children from two-parent families were
more likely to have known family history of coronary heart disease (CHD) before 60 years of age (41.8% vs
25.8%, P < .001), and twice as likely as children from
single-parent families to have known parental hypercho-lesterolemia (18.8% vs 9.5%, P < .001). Only 39% of par-ents reported having had their cholesterol level checked; they were better educated and more likely to have health insurance. Parents with a first-degree relative with CHD before 60 years of age were more likely to report having
their cholesterol level checked and to report a high
cho-lesterol level. We calculated that 27% of children (18% of
children from single-parent households and 29% of
chil-dren from two-parent households) would report a known
family history of premature CHD (ie, CHD before 55
years of age) and qualify for lipoprotein analysis, and that 11% of children would qualify for total cholesterol screening because of known parental hypercholesterol-emia without a family history of premature CHD.
Thirty-five percent of children had incomplete or unavailable family health history and/or unknown parental choles-terol status.
Conclusions. In this population, 38% of children
would be targeted for cholesterol screening, exceeding the estimate of the NCEP Children and Adolescents Panel. The selection process, however, would tend to miss children from single-parent families, children with incomplete family health history, and children whose
parents have not had their cholesterol levels measured.
The currently recommended pediatric cholesterol
screen-From the The Mary Imogene Bassett Research Institute, Bassett Healthcare,
Cooperstown, NY.
Presented, in part, at the 64th Annual Meeting of the American Heart
Association, Anaheim, CA, November 1, 1991. Abstract published in
Circu-lation. 1991; 84(Suppl); 11-33.
Received for publication May 19, 1993; accepted Jan 27, 1994. Reprint requests to (BAD.) The Mary Imogene Bassett Research Institute,
Bassett Healthcare, One Atwell Road, Cooperstown, NY 13326-1394.
PEDIATRICS (ISSN 0031 4005). Copyright © 1994 by the American
Acad-emy of Pediatrics.
ing policy needs to be evaluated further in additional communities and population settings. Alternative choles-terol screening strategies are needed when family health
history is incomplete and/or parental cholesterol status
is unknown. Pediatrics 1994;94:296-302; cholesterol, screening, childhood.
ABBREVIATIONS. NCEP, National Cholesterol Education
Pro-gram; LDL, low-density lipoprotein; CHD, coronary heart
dis-ease; LRC, Lipid Research Clinics; HMO, health maintenance organization.
There is little doubt that elevated blood choles-terol, specifically elevated low-density lipoprotein (LDL) cholesterol, is an important risk factor for coronary heart disease (CHD).’3 In adults, lowering elevated cholesterol levels is associated with a
reduc-tion in CHD events and death.46 Autopsy studies
reveal that the earliest stages of atherosclerosis begin asymptomatically during childhood and adolescence and are related to cholesterol and lipoprotein levels, blood pressure, and cigarette smoking.79 The Adult
Treatment Panel of the National Cholesterol
Educa-lion Program (NCEP) has recommended that all
adults >20 years of age be screened with a total cholesterol measurement to detect those with ele-vated LDL-cholesterol levels.’0 With few exceptions
these guidelines have generally been accepted,
though implementation into practice has been
achieved to a lesser extent.
The question of cholesterol screening in children has raised considerable, and quite heated, controver-sy.”6 In April 1991, the NCEP Expert Panel on Blood Cholesterol Levels in Children and Adolescents released
their recommendations for cholesterol screening,17 which
were widely disseminated in March 1992, and supported
by the American Academy of Pediatrics, Committee on
Nutrition in September 1992.18 The NCEP Children’s
Panel recommends that: children and adolescents 2
years of age with a family history (parent or grandparent)
ofpremature cardiovascular disease (coronary
atherosde-rosis, myocardial infarction, angina pectoris, peripheral
vascular disease, cerebrovascular disease, or sudden
car-diac death, with onset on or before 55 years of age) should be tested for dyslipoproteinemia with a lipoprotein anal-ysis; and children and adolescents 2 years of age, whose parent has had hypercholesterolemia (total cholesterol value >621 mmol/L or 244) mg/dL), but who do not have a family history of premature cardiovascular disease,
should be screened for elevated LDL-cholesterol by
men-suring a total cholesterol level. The NCEP Children and
children would have a positive family history of
prema-ture cardiovascular disease and qualify for cholesterol
screening with lipoprotein analysis They estimated that
an additional 19.5% would have parental
hypercholester-olemia alone and require a total cholesterol measurement
(they predicted that 25.1% of children would have paren-tal hypercholesterolemia, and of these they expected 5.6% would also have a positive family history; therefore 25.1%
- 5.6% = 19.5%). However, the 5.6% estimate was based
on the prevalence of premature CHD in children’s par-ents in the Upid Research Clinics (LRC) Population Stud-ies.17 Data were not available regarding prevalence of parental history ofother cardiovascular diseases nor prey-alence of premature cardiovascular disease in the
chil-chen’s grandparents
The purpose of this study was to determine, from a well-defined total population sample, the number and percent of children and adolescents with known
and reported family history of premature CHD and
known and reported parental hypercholesterolemia, who would be targeted for cholesterol screening
un-der the current NCEP Children and Adolescents
Panel Guidelines.
Population
METHODS
Health Census ‘89 was a survey of all households in Otsego
County, an Upstate New York county composed of a predomi-nantly rural (90%), white (98%), lower-middle to middle class population.19 A team of cartographers mapped every dwelling
unit in the entire county, an area of 1000 square miles. A
ques-tionnaire was mailed to each household, and the completed ques-tionnaires were collected by trained field staff. Only permanent residents of Otsego County were eligible; those residing less than 6 months in Otsego County or those who did not consider their Otsego County residence to be their permanent residence were
excluded. Of 19 800 permanent households identified, 17 444
households successfully completed the questionnaire (86.6%
re-sponse rate). Of the remaining, 2.2% did not provide useful
infor-marion, 2.9% could not be contacted, and 8.3% refused to
partic-ipate. There were 44 565 participants, of whom 10 457 were
children, aged 2 through 19 years, living with at least one parent
(Table 1).
Questionnaire Data
Data on all adults aged 20 years were collected by self-report. Information collected on the children’s parents included:
demo-graphic information; frequency of health visits; health insurance
status; cigarette smoking status; and medical history, including high blood pressure, CHD diagnosed by a physician, early CHD
(onset before 60 years of age) in a first-degree relative (sibling
and/or parent), whether blood cholesterol had been measured in
the past 2 years, and history of “high cholesterol” diagnosed by a physician. “High cholesterol” levels were not further defined.
Statistical Analysis
All data were dual-entered and verified. Chi-square (‘) tests
were used for dichotomous variables and Student’s t tests for continuous data. The relationship between prevalence of known
parental hypercholesterolemia and known family history (parent,
grandparent, or blood-related aunt or uncle) of CHD before the
age of 60 years versus the age of the child (in 3-year age groups)
was examined by the Mantel-Haenszel chi-square test. Using our
population-based data we calculated the increase in prevalence of
doctor-diagnosed CHD between 55 and 60 years of age. We made
a crude adjustment downward to obtain a conservative estimate of
the number of children with a family history of “premature” CHD (ie, CHD with onset before age 55 years in a parent or grandpar-ent) (Appendix). Proportionate changes were also made to adjust
the prevalence of family history of premature heart disease among
those with known parental hypercholesterolemia. All statistical tests were two-sided. Because of multiple comparisons, P tests < .01 were considered statistically significant. All analyses were
conducted using SAS (Version 606, SAS Institute Inc., Cary, NC).
RESULTS
The children were equally distributed by age (2 through 19 years) and sex (Table 1). Sixteen percent (1677) lived with only one parent (84% with moth-ers); the remainder (8780) lived with both parents. The percent of children with known family history (parent, grandparent, or blood-related aunt or uncle) of CHD before 60 years of age and the percent with
known parental hypercholesterolemia increased
with the age of the child (both P < .001, Mantel-Haenszel
x
= 80 and 211, respectively) (Fig 1). Mostof the family history of early CHD was not due to
CHD in the parents. Only 5.8% of children had
par-ents with CHD: 2.2% of children’s mothers, 3.8% of
children’s fathers, and in 0.3% of children, both
par-ents had CHD. Reported hypertension also increased
with parental age; 7% of women and 14% of men,
aged 20 to 40 years, reported elevated blood
pres-sure, while 22% of women and 26% of men, aged 40
to 60 years, reported high blood pressure. Smoking rates did not vary with parental age; 29% of mothers
and 31 % of fathers reported currently smoking
cigarettes.
Children living in two-parent households were
more likely to report a known family history of early CHD than children living in single-parent house-holds (41.8% vs 25.8%, P < .001, = 151) (Fig 2).
Children from two-parent families were twice as
likely as children from single-parent famifies to have
known parental hypercholesterolemia (18.8% vs
9.5%, P < .001, = 85). Because family health history
from both sides of the family is more likely to be
known when children live with both parents, a
TABLE 1. Study Population (N = 10 457) Age, Sex, and Family Status
Age (Years) Male Female Total
Single-Parent Two-Parent Single-Parent Two-Parent
2 through 4 129 797 137 765
5through 7 167 818 146 798
8 through 10 144 764 139 749
II through 13 143 748 141 652
14 through 16 113 698 140 608
17 through 19 141 753 134 617
1828 1929 1796
1684 1559
1645
Total 837 4578 837 4189 10 441*
50
40
30
20
10
0 C’..
0
0 5)
0.
2-4 5-7 8-10 11-1J 14-1
Single-Parent Families (N 1 677) Two-Parent Families (N=8780)
5.1% 4.4%
I:i
.
21.4%.,
..
31.2%10.6%:
8 27 0
Age (years)
Parent High Cholesterol ::Fomily History of CHD
Fig 1. The percent of children with known family history (parent,
grandparent, or blood-related aunt or uncle) of CHD before 60 years of age and the percent of children with known parental hypercholesterolemia increased with the age of the child (both P <
.001, Mantel-Haenszel = 80 and 211, respectively).
69.17, 50.07,
LI1]FH Only Both HC Only Neither
Fig 2. Children living in two-parent families were more likely
than children from single-parent families to have known family
history (parent, grandparent, or blood-related aunt or uncle) of
CHD before 60 years of age (FH Only and Both) and known parental hypercholesterolemia (HC Only and Both) (both P< .001,
x2 = 151 and 85, respectively).
higher percentage of children living in two-parent
families would be identified with a risk factor for
premature CHD and qualify for cholesterol
screen-ing compared to children living in single-parent
families (50% vs 31%, P < .001,
x2
= 205).Only a small percent of the children who had a
positive family history also had known parental
hy-percholesterolemia: 17% of children in single-parent
families (N = 74, 4.4% of all children in single-parent
families); and 25% of children in two-parent families
(N = 933, 10.6% of all children in two-parent
fami-lies). This is probably due, in part, to the fact that
only 39% of parents (7549) reported having had their
cholesterol levels measured during the past two
years; 1812 (24.0% of those tested) reported being
told that they had a “high” cholesterol level. Parents
with a first-degree relative with CHD before age 60
years were more likely to have had their own
cho-lesterol levels checked (46% vs 39% of single parents,
respectively, P < .005,
x
= 8; among two-parentfamilies with a first-degree relative with early CHD,
one parent was tested in 34% and both parents were
tested in 30%, compared to families without a
first-degree relative with CHD, where one parent was
tested in 31 % and both parents were tested in 22%, P
< .001,
x2
= 112) (Table 2). Parents with a first-degreerelative with CHD were also more likely to report
high cholesterol levels (37% vs 19% of single parents;
and in two-parent families, 34% vs 22% if one
par-ent’s cholesterol level was checked, and 47% vs 37%
if both parents’ cholesterol levels were checked, all P
< .001, x2 25, 51, 21, respectively) (Table 2). There
was concordance in cholesterol testing by parents
living together. If one parent had his/her cholesterol
level measured, the other was also more likely to
have had his/her cholesterol level checked. In 42% of
two-parent families, neither parent had their choles-terol level checked; in 26%, both parents had their
cholesterol checked; in 16%, only the father’s
choles-TABLE 2. Relationship Between Whether Parents Had Their Cholesterol Levels Checked and High Parent Cholesterol Versus History
of CHD Before Age 60 Years in a First-Degree Relative
Number of parents Single-Parent Families Number of Two-Parent Families
who had cholesterol (N 1677) parents who had (N 8780)
levels checked Relative With CHD cholesterol levels First-Degree Relative with CHD
First-Degree checked
No Yes No Yes
N (%) N (%) N (%) N (%)
0 716 (61) 228 (54) 0 2159 (46) 1248 (35)
I 455 (39) 192 (46) 1 1461 (31) 1214 (34)
1171 (100) 420(100) 2 1042 (22) 1068 (30)
(P < .005, x2= 7.8) 4662 (100) 3530 (100)
(P < .001, x2 112.4)
High Single-Parent Families Two-Parent Families
cholesterollevel (Cholesterol Checked in I Parent) (Cholesterol Checked in 1 Parent) (Cholesterol Checked in 2 Parents)
p . First-Degree First-Degree First-Degree
Relative With CHD Relative With CHD Relative With CHD
No (%) Yes (%) No (%) Yes (%) No (%) Yes (%)
No 369 (81) 125 (63) 1143 (78) 800 (66) 656 (63) 567 (53)
Yes 86 (19) 74 (37) 318 (22) 414 (34) 386 (37) 501(47)
455 (100) 199 (100) 1461 (100) 1214 (100) 1042 (100) 1068 (100)
terol was checked; and in 17%, only the mother’s cholesterol was checked (P < .001, = 899).
Children, living in either single-parent or two-parent families whose parents reported having their cholesterol level checked, had better educated
par-ents (13.3 vs 12.7 years and 14.0 vs 13.1 years of
schooling, respectively, both P < .001). Children
liv-ing in two-parent families, whose parents had their cholesterol level checked, lived closer to health care
(9 vs 10 miles, P < .001) and reported using the health
care system more often (1 1 vs 9 times per year, P <
.001). Health insurance status was also significantly
related to parental cholesterol testing. Parents who
belonged to a health maintenance organization
(HMO) were much more likely to have had their
cholesterol levels measured than parents with
private group insurance, Medicaid, or no health
insurance (Table 3).
Because our study defined “early” CHD as CHD
with onset prior to 60 years of age rather than prior to 55 years of age as suggested by the NCEP Chil-dren and Adolescents Panel, we made an adjustment to calculate the percent of children expected to have
a family history of “premature” CHD (ie, CHD with
onset before 55 years of age) (see Appendix). This
resulted in a decreased estimate of the percent of
children qualifying for cholesterol screening. Twen-ty-nine percent of children from two-parent families
(reduced from 41.8%) and 18.4% of children from
one-parent households (reduced from 25.8%) would qualify for lipoprotein analysis (Table 4). Eleven per-cent of children from two-parent families and 6.3% from single-parent families would qualify for total cholesterol screening. Overall in this population, 38% of children would qualify for cholesterol screening:
27% a lipoprotein profile and 11 % a total cholesterol
measurement. An additional 35% of children (28% of children in two-parent families, and the remaining 69% of children in single-parent families) had incom-plete or unavailable family health history and/or unknown parental cholesterol status.
DISCUSSION
We found essentially the same rate of premature
CHD in parents (5.8%) as reported by the NCEP
Expert Panel on Blood Cholesterol Levels in Children
and Adolescents based on the LRC Population Study for 1- to 19-year-olds (5.6%), even though the parents in our study were a few years younger (mother’s
mean age = 37 years and father’s mean age = 40
years) than in the LRC Population Study (mean
par-ent age = 41 years).’7 When family history
ascertain-ment included the children’s aunts, uncles, and
grandparents, many more children had a positive
family history of premature CHD. Since the preva-lence of CHD increases with age, and grandparents are older than parents, it is not surprising that more
grandparents than parents have experienced CHD.
In this population, 39% of children had a family
history of CHD before 60 years of age. After
adjust-ment, we calculated that at least 27% of children
(29% from two-parent families and 18% from
single-parent families) would have a parent and/or
grand-parent with premature CHD. Our calculation that
grandparents would be responsible for a positive
family history of CHD about four times as often as
parents, is the same as reported by Williams and
colleagues.2#{176} r estimate that 27% of children have
a family history of premature CHD, though higher
than the 5.6% estimate made by the NCEP Childrens
Panel,’7 is comparable to the 31 %, 26%, and 35%
reported in studies of children from Illinois, Texas, and Utah, respecfively.20’’ It is also similar to the
estimate calculated from data reported in several
previous studies. By combining the finding that
chit-dren whose parents or grandparents experienced
early CHD were 2 to 2.5 times more likely to have
hyperlipidemia (elevated total and/or
LDL-choles-terol levels) than children without such a history,27 with the finding that 50% of children with an ele-vated total and/or LDL-cholesterol levels had a pos-itive family history of premature CHD,1121’2 we
cal-culate that 33% and 29%, respectively, of the
pediatric population would be expected to have a
positive family history of premature CHD.
There may be several reasons why family history of premature cardiovascular disease identifies only half the children with hypercholesterolemia. Young
children tend to have young parents and young
grandparents, who may be too young to have
expe-rienced premature CHD. In addition, a significant number of children have incomplete or unavailable
TABLE 3. Relationship Between Whether Parents Had Their Cholesterol Level Checked Versus
Insurance Status
Insurance Status Single-Par ent Families Two-Pare nt Families
Children,
N
Parent Cholesterol
Checked,
N(%)*
Children, N
Parent Cholesterol
Checked,
N(%)*
Private group/employer Medicaid
None (self-pay)
Health maintenance organization
765 370 108 130
321 (42) 148 (40) 40 (37) 74 (57)
5344 419 559 987
3206 (60) 155 (37) 302 (54) 790 (80)
Total 1373 538 (42) 7309 4453 (61)
* Number and percent of children with at least one parent who has had his/her cholesterol level
checked.
TABLE 4. Children Qualifying for Cholesterol Screening
Risk Factor Present Single-Parent Families
(%)
Two-Parent Families
(%)
All Children
(%)
Unadjusted Adjusted Unadjusted Adjusted* Unadjusted Adjusted*
Family history of early CHD Parent
Other relative*
3.5 3.5
25.0 14.8
5.1 5.1
39.2 23.9
4.8 36.9
4.8 22.4
Parent and/or other relative 25.8 18.4 41.8 29.0 39.2 27.2
Parent hypercholesterolemia With family history of CHD Without family history of CHD
4.4 3.1
5.1 6.3
10.6 7.4
8.2 11.4
9.6 7.7
6.7 10.6
With or without family history 9.5 9.5 18.8 18.8 17.3 17.3
Family history of CHD and/or parent hypercholesterolemia 31 25 50 40 47 38
* Unadjusted “other relative” family history includes grandparents and blood-related aunts and uncles with CHD before age 60 years. Adjustment was made to limit family history of CHD to CHD beginning before age 55 years.
family health history due to an absent or unknown
parent and/or grandparent. In the United States,
26% of children are born to single mothers: 18% of
white children, 64% of black children, and 34% of Hispanic children.29 Twenty-five percent of children live with only one parent; 88% of these children are raised by mothers (84% in this study). In many cases, the father’s (or a generation previous, the grandfa-ther’s) health history may not be known. Because of divorce and separation, 50% of children, before the age of 18 years, wifi live in a single-parent house-hold. The impact of family disintegration on one’s ability to ascertain a complete family health history is not known, but it is almost certainly adverse.
In this study, the questions were asked in such a way as to allow separate analyses of children based on whether family history data was available from one or two parents. As one would expect, children
who had family health history data from only one
parent were less likely to have a known family his-tory of CHD and less likely to have known parental hypercholesterolemia than children who had family health history data reported from both parents. Thus, the children missed by a selection policy based on family health history are not random; they are likely to be systematically missed because of the inability to
obtain a complete family health history. The NCEP
Children and Adolescents Panel did not specifically
address this issue, but did state that in circumstances where family history is not available, the physician “may choose” to measure cholesterol levels.’7 Unless the family history data is collected in a sensitive, yet
systematic way, however, children with an
incom-plete or unavailable family health history will be indistinguishable from those with a negative family health history. In many published studies, this dis-tinction has not been made, and in clinical practice,
the questions are often not asked in a manner to
make the distinction.
Another problem not adequately addressed by the
NCEP Children and Adolescents Panel is that many
parents have not had their cholesterol levels mea-sured and, therefore, cannot know if they are ele-vated. In this study, only 39% of parents reported having their cholesterol levels checked during the 2 years prior to the survey (1987 to 1989). Because the
NCEP Adult Treatment Panel Guidelines were not
issued until 1988,10 one would expect that awareness
of cholesterol levels and testing would have
in-creased after 1988 so that most adults tested were
probably tested recently. Our finding, that in 1989,
39% of parents reported having had their cholesterol level tested in the past 2 years is consistent with national data from 1989, where 37% of 18- to 34-year-olds and 61 % of 35- to 49-year-olds reported ever having had their cholesterol level checked,#{176} and an-other study, where 44% of parents of elementary school children reported having their cholesterol checked in the past year.3’ Because of increased cho-lesterol awareness, one would expect that, with time, more parents will have their cholesterol levels mea-sured. Furthermore, with the emphasis on “managed competition,” it is expected that more individuals wifi participate in HMOs. Since individuals enrolled in HMOs were more likely to have their cholesterol level measured, one would anticipate with the shift to HMOs that more adults will have their cholesterol levels measured and that more children will qualify for cholesterol screening in the future because of known parental hypercholesterolemia.
In this study, 24% of parents who had had their cholesterol levels measured reported that they were “high.” Overall, 17% of children reported at least one parent with “high cholesterol.” Because of the self-selection bias (ie, parents with a first-degree relative with early CHD were both more likely to have their
cholesterol level measured and to report it was
“high”), one cannot extrapolate this percentage to parents who have not had their cholesterol levels
measured. In addition, the parents’ perception or
their health care provider’s definition of “high” cho-lesterol may not coincide with the total cholesterol cutpoint of 240 mg/dL chosen by the NCEP Children and Adolescents Panel. Asking parents for their cho-lesterol levels is also problematic. A recently pub-lished study reported that only two-thirds of parents
who reported having had their cholesterol levels
measured in the past year reported a known choles-terol level.31
Better educated individuals and those with better
access to health care (financial and physical) are
mea-sured.#{176}32 Children living in single-parent house-holds are the poorest segment of the U.S. population, with 64% of the 12.8 miffion children who are below the poverty level residing in single-parent house-holds.29 The education level of the head of household
is inversely related to the proportion who live in
poverty. Less educated individuals also have higher mortality rates, and as a group have made the least
improvement in CHD mortality? In addition,
chil-dren from these low income families and children with poorly educated heads of household consume more atherogenic diets (ie, consume diets with more calories from total fat, saturated fatty acids, and di-etary cholesterol) than children with better-educated parents earning higher incomes.M Consequently, poor children and children who live in single-parent households constitute a population at increased risk
for premature CHD. However, because of
incom-plete or unavailable family health history and/or unknown parental cholesterol status, without special efforts and alternative screening methods, these chil-dren will tend to be missed by the current pediatric cholesterol screening strategy.
CONCLUSION
The NCEP Expert Panel on Blood Cholesterol
Lev-els in Children and Adolescents recommends a
se-lective cholesterol screening strategy based on the child’s family health history and the parents’ choles-terol status. Premature CHD and hypercholesterol-emia cluster in families, and a selective screening strategy takes advantage of this clustering. More-over, a child with hypercholesterolemia and a posi-tive family history of premature CHD is more likely to remain hypercholesterolemic as he grows older,
and probably more likely to develop premature
CHD than a child with the same cholesterol level and
a negative family history of premature CHD. But
nothing can be said about the child with hypercho-lesterolemia and an unknown or unavailable family history of premature CHD.
Ascertaining a complete family health history is
time-consuming2#{176} and often difficult, as was noted
by the Panel,’7 commented on by Dr Lauer,37 and
indicated by the absence of reliable estimates for the prevalence of premature CHD in children’s grand-parents. Furthermore, because of the disintegration of the American family, for many children, informa-tion regarding their family health history and/or parent cholesterol status will simply not be available. In this study, one-third of children had incomplete
family health history and/or unavailable or
un-known parent cholesterol status. In some pediatric
populations these children may predominate; in a
recent study from a low-income, inner city practice,
76% of children had incomplete family health
histo-ries. These high risk children and their families
need to be identified and targeted, not only for
cholesterol screening, but for primary CHD risk
reduction.
In our study population, we calculated that 38% of children would be targeted for cholesterol screening
based on known family history of premature CHD
and/or parental hypercholesterolemia. This exceeds
the estimate of the NCEP Children and Adolescent Panel, primarily because we included grandparents
with premature CHD. Consequently, the costs of
implementating the pediatric cholesterol screening
guidelines will be significantly more than was
previously estimated.’7’9
Cholesterol screening, however, is only the first step in identifying those with hypercholesterolemia. Benefits will accrue only by treating those identified. Research is needed to evaluate in different pediatric
settings, the implementation, efficacy, and costs of
the selective cholesterol screening recommendations compared to alternative cholesterol screening
strate-gies. In addition, pediatric cholesterol treatment
pro-tocols that are culturally sensitive, economically fea-sible, safe, and practical for children, families, and
pediatric care providers need to be developed and
evaluated in the diverse pediatric practice settings and communities across this country.
APPENDIX
We made a conservative adjustment to reduce the estimate of the number of children expected to have a “family history of premature CHD” to account for the different definitions used in
our data base compared to that adopted by the NCEP Expert Panel on Blood Cholesterol Levels in Children and Adolescents.
The proportions of women aged 54 and <55 years and 59 and <60 years reporting doctor-diagnosed CHD were 8.25% and 8.76%, respectively. The proportions of men, aged 54 and <55 years and 59 and <60 years reporting doctor-diagnosed CHD were 10.18% and 19.10%, respectively. We created a crude adjust-ment factor to reduce the estimate of the number of children with a family history of premature CHD, calculated as the proportion of CHD occurring in men before age 60 years multiplied by the ratio of CHD rate in men 54 and <55 years to CHD rate in men 59 and <60 years. A similar calculation was made for women and the two proportions were added:
0.1018 0.1910 0.0825 0.0876
x + x = 0.661.
0.1910 0.2786 0.0876 0.2786
The family history contribution of the children’s second-degree blood relatives (ie, the parent’s first-degree relatives) was reduced by multiplying by this fraction, separately for each 3-year age group of children, separately for children living in one-parent or two-parent families.
Children whose parents were 55 years of age were deleted
(N = 440). No additional adjustment was made for the potential
contribution to family history by blood-related aunts uncles for several reasons. One would expect the CHD prevalence rate in the children’s aunts uncles to equal approximately the rates in the mothers fathers (2.2% and 3.4%, respectively). One would also expect some overlap (or concordance in history of early CHD) between the parents and their siblings (ie, the children’s aunts and uncles) and their parents (ie, the children’s grandparents). Thus the reduction would probably be a fraction of the rates in the mothers and fathers. These percentages are small compared to the 34% reduction already made.
We consider this reduction conservative, because it assumes that all the grandparents are 60 years of age and have, therefore, experienced the total increased probabifity of CHD occurring be-tween the ages of 55 and 60 years. Second, our family history data included only heart disease (myocardial infarction, heart attack, bypass surgery, angina pectoris, sudden cardiac death). Peripheral vascular disease and cerebrovascular disease were not included. Therefore, our estimate is further underestimated by the prevalence rates of these conditions.
ACKNOWLEDGMENTS
REFERENCES
1. Casteffi WP, Garrison RJ, Wilson PWF, et a!. Incidence of coronary heart disease and lipoprotein cholesterollevels. The Framingham Study. JAm Med Assoc. 1986256:2835-2838
2. Kagan A, McGee DL, Yano K, et al. Serum cholesterol and mortality in
a Japanese-American population: The Honolulu Heart Program. Am J
Epidemiol. 1981;114:11-20
3. Stamler J, Wentworth D, Neaton JD. Is relationship between serum
cholesterol and risk of premature death from coronary heart disease continuous and graded? Findings in 356,222 primary screenees of the Multiple Risk Factor Intervention Trial (MRFIT). JAm Med Assoc. 1986;
256:2823-2828
4. Lipid Research Clinics Program. The Lipid Research Clinics Coronary Primary Prevention Trial Results. II. The relationship of reduction in incidence of coronary heart disease to cholesterol lowering. JAm Med i4ssoc. 1984251:365-374
5. Frick MH, Elo 0, Haapa K, et al. Helsinki Heart Study:
primary-prevention trial with gemfibrozil in middle-aged men with
dyslipi-demia. Safety of treatment, changes in risk factors, incidence of coro-nary heart disease. N Engi JMed. 1987317:1237-1245
6. Carison LA, Rosenhamer G. Reduction of mortality in the Stockholm
Ischaemic Heart Disease Secondary Prevention Study by combined
treatment with clofibrate and nicotinic acid. Acta Med Scand. 1988;223: 405-418
7. Newman WP ifi, Freedman DS, Voors AW, et al. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engi JMed. 1986314:138-144
8. Newman WP ifi, Wattigney W, Berenson CS. Autopsy studies in U. S. children and adolescents. Relationship of risk factors to artherosclerotic lesions. Ann NY Aced Sd. 1991;623:16-25
9. PDAY Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking. A prelimi-nary report from the pathobiological determinants of atherosclerosis in youth (PDAY) research group. JAm Med Assoc. 1990264:3018-3024 10. National Cholesterol Education Program. Report of the Expert Panel on
Detection, Evaluation, Treatment of High Blood Cholesterol in Adults. Be-thesda, MD#{149}.U. S. Department of Health and Human Services, Public Health Service, Nationallnstitutes of Health, National Heart, Lung, and Blood Institute, January 1989; NIH Publication 88-2925.
11. Garcia RE, Moodie DS. Routine cholesterol surveillance in childhood.
Pediatrics. 1989;84:751-755
12. Feldman W. Routine cholesterol surveillance in childhood. Pediatrics.
199086:150 [letter].
13. Wynder EL, Berenson CL, Strong WB, Wffliams C (eds). Coronary artery disease prevention: cholesterol, a pediatric perspective. Prey Med.
1989;18:323-409
14. Lauer RM, Clarke WR. Use of cholesterol measurements in childhood for the prediction of adult hypercholesterolemia. The Muscatine Study. JAm Med Max. 1990;264:3034-3038
15. Newman TB, Browner WS, Hulley SB. The case against childhood cholesterol screening. JAm Med Assoc. 1990;264:3039-3043
16. Resmcow K, Berenson C, Shea 5, et al. The case against “the case against childhood cholesterol screening”. JAm Med Assoc. 1991;265:3003-3005
[comment]
17. National Cholesterol Education Program. Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents. Bethesda, MD: U. S. Department of Health and Human Services, Public Health Service, National Institutes of Health, NationaiHeart, Lung, and Blood Institute, September 1991; NIH Publication 91-2732.
18. American Academy of Pediatrics, Committee on Nutrition. Statement on cholesterol. Pediatrics. 1992;90:469-473
19. Otsego County Planning Department. Otsego County Housing Market
Study. Cooperstown, NY: Otsego County Planning Department; 1987
20. Wffliams RR, Hunt SC, Barlow GK, et al. Prevention of familial
cardio-vascular disease by screening for family history and lipids in youth. Clin
Chem. 1992;38:1555-1566
21. Griffin TC, Christoffel KK, Binns H, et al. Family history evaluation as a predictive screen for childhood hypercholesterolemia. Pediatrics. 1989; 84:365-373
22. Nora ll Lortscher RH, Spangler RD. et al. Genetic epidemiologic study of early onset ischemic heart disease. Circulation. 1980;61:503-508
23. Glueck CJ, Fallat RW, Tsang R, et al. Hyperlipidemia in progeny of parents with myocardial infarction before age 50. Am JDis Child. 1974;
127:70-75
24. LeeJ, Lauer RM, Clarke WR. Lipoproteins in the progeny of young men with coronary artery disease: children with increased risk. Pediatrics.
1986;78:330-337
25. Schrott HG, Clarke WR, Wiebe DA, et al. Increased coronary mortality in relatives of hypercholesterolemic school children: The Muscatine Study. Circulation. 1979;59:320-326
26. Moll PP, Sing CF. Weidman WH, et al. Total cholesterol and lipopro-teins in school children: prediction of coronary heart disease in adult relatives. Circulation. 1983;67:127-134
27. Dennison BA, Kikuchi DA, Srinivasan SR, et al. Parental history of
cardiovascular disease as an indication for screening for lipoprotein abnormalities in children. JPediatr. 1989;115:186-194
28. Shea 5, Basch CE, Ingoyan M, et al. Failure of family history to predict high blood cholesterol among Hispanic preschool children. Prey Med.
1990;19:443-455
29. U. S. Bureau of the Census, Statistical Abstract of the United States: 1991.
111th ed, Washington, DC: U. S. Government Printing Office; 1991. 30. Factors related to cholesterol screening cholesterol level
awareness-United States, 1989. Morb Mortal Weekly Rep. 199039:633-637
31. Resnicow K, Cross D. Are parents’ self-reported total cholesterol levels useful in identifying children with hyperlipidemia? An examination of current guidelines. Pediatrics. 1993;92:347-354
32. Rybicki B, Pearson TA. Who gets their cholesterol measured? Predictors
in a rural population. Circulation. 1990;82(Suppl):llI-515 [abstract].
33. Feldman ll Makuc DM, Kleinman JC, Cornoni-Huntley J. National
trends in educational differentials in mortality. Am JEpidemiol. 1989; 129:919-933
34. Martinez GA, Ryan AS. What are children in the United States eating?
In: Roche AF, ed, Prevention of Adult Atherosclerosis During Childhood,
Report of the Ninety-Fifth Ross Conference on Pediatric Research. Columbus, OH: Ross Laboratories; 198853-60
35. Thompson FE, Dennison BA. Dietary sources of fats and cholesterol in USchildren aged 2 through 5 years. Am JPublic Health. 1994;84:799-806 36. Freedman OS, Shear CL, Srinivasan SR. et al. Tracking of serum lipids
and lipoproteins in children over an 8-year period: The Bogalusa Heart
Study. Prey Med. 1985;14:203-216
37. Lauer RM. Should children, parents, and pediatricians worry about
cholesterol?. Pediatrics. 1992;89:509-511 [comment]
38. Wadowski SJ, Karp RJ, Murray-Bachmann R, Senft C. Family history of
coronary artery disease and cholesterol: Screening children in a disad-vantaged inner-city population. Pediatrics. 199393:109-113
