Usefulness of Parental Serum Total Cholesterol Levels in Identifying Children with Hypercholesterolemia Irwin Benuck, MD, PhD, Samuel S. Gidding, MD, Mark Donovan, Edward S. Traisman, MD, and Howard S. Traisman, MD
It was hypothesized that healthy children wlth high cholesterol levels may have parents who exceed acceptabte cholesterol levels established by the National Chotesterol Education Program. One hundred slxty families (320 parents, 263 children aged 3 to 10 years) were evaluated for total cholesterol and other rlslc factors. Defore the study, almost half of the parents had net had serum total cholesterol measured. The odds ratlo for a child having a total chotesterol 25.17 mmoi/liter (266 mg/dl) was 13.&l (confidence interval 5.7 to 32.S) for a child with at teast 1 parent having cholesterol 26.26 mmol/Mter (246 mg/dl) versus a chitd whose parents had’low total cholesterol. Testing only chtldren who had at least 1 parent wlth a total chotesterol 25.17 mmol/liter (260 mg/dl) had a sensltlvity of 98% for detecting children’s total chotesterol1S.17 mmol/liter. It is concluded that parental total cholesterol is useful in ldentlfying children with high total cholesterol levels. Pediatricians may identify a large number of parents with hypercholesterolemia not previously recognized. (AmJCardid 1992;69:713-717)
hildren with elevated total cholesterol levels are likely to have elevated total cholesterol levels as adults.1-5In an effort to establish guidelines for identifying hypercholesterolemia in children, the National Cholesterol Education Program (NCEP) pediatric panel has suggested that children with a family history of heart disease,family history of hypercholesterolemia or parental serum total cholesterol 16.20 mmol/liter (240 mg/dl) be evaluatecL6Several studies have challenged this statement as being too restrictive, concluding that many children with hypercholesterolemia would be excluded from screening.7~8 Today, there is a lack of uniform strategy among pediatricians to identify children at risk for elevated cholesterol as adults?*iO To determine children at risk for elevatedcholesterol in a private pediatric office, we evaluated a variety of familial risk factors including parents’ serum total cholesterol to identify preadolescentchildren who should be evaluated for hypercholesterolemia. We hypothesized that children with high cholesterol would have parents who exceedacceptablecholesterol concentration established by the NCEP.” In addition, we hypothesized that comprehensivefamily screeningrather than assessment of individual children’s total cholesterol would be beneficial in identifying children and parents with hypercholesterolemia.
C
METHODS
Subjeetsr We asked 180 biologic families to enter the study, and completedata were obtained on 160 families with 263 children between the ages of 3 and 10 years. Data from these 160 families are included in this report. All were Caucasian.Children were identified at routine pediatric visits. Only parents and children who From the Children’s Memorial Hospital, Northwestern University were in good health were allowed entry. Pregnant womMedical School,Chicago, Illinois. This study was supportedin part by en or any subjectswith chronic illness, diabetes,renal or the Frieda Stone Heart Fund, Highland.Park, Illinois, and the Chilliver diseasewere excluded from participation because dren’s Memorial Institute for Researchand Education, Chicago, Illinois.Dr. Gidding is a memberof the FeinbergCardiovascularResearch of possible effects of these conditions on lipid levels.‘* Institute, Chicago, Illinois. ‘Manuscript received Septembci 3, 1991; Participants were recruited between June and Novemrevisedmanuscript receivedNovember 22,1991, and acceptedNovem- ber 1989 from a single group pediatric practice. ber 30. wre: A questionnaire was completed by Addressfor reprints: SamuelS. Gidding, MD, Division of Cardiology #21, Children’s Memorial Hospital, 2300 Children’s Plaza, Chica- both parents which asked about personal and family risk factors. Information was sought regarding prior go, Illinois 60614. PEDIATRIC CHOLESTEROL SCREENING
713
TABLE I Demographic Information on Participating Families Father No. of subjects
Age (yr) (range) Total cholesterol (mmol/liter) (range) Body mass index (kg/m*) (range) Family history of CAD (%) Family history of elevated total cholesterol (%) Tobacco use (%I No prior TC screen (%)
Mother
160 38 A 5 (27-61) 5.37 -c 1.8 (3.54-8.48) 25 r 3 (18-37) 48.8 36.6
160 36 r 4 (25-47) 5.01 k 0.9 (2.53-7.73) 23 f 4 (17-37) 51 33.1
11.9 43
Children 263 5.7 + 4.50 k 15.8 +
(149M/114F) 2.1 (3-10) 0.8 (2.71-8.40) 2 (10.4-25.3) 71 53
8.1 54
Values are mean + standard deviation. CAD = coronary artery disease; TC = total cholesterol. I
cholesterol testing of each parent, parental smoking, family history of coronary artery disease(angina, myocardial infarction, coronary bypass surgery or sudden cardiac death) and family history of elevatedtotal cholesterol or use of a lipid-lowering medication. A positive family history of coronary artery diseasewas defined as a positive responsefor that parent or a first-degree relative of either parent. A positive family history of elevated total cholesterol was defined as a report of total cholesterol 15.17 mmol/liter (200 mg/dl) in the parent or a first-degree relative of the parent. Clink~I aueumant: All subjectswere weighed and measuredusing office scales.Body mass index (weight [kg1/heightb211 was calculated for each subject.13 Nonfasting blood specimenswere obtained by finger stick. Freely flowing capillary blood was collected in a commercially available collection tube (Norfolk Scientific Collection Tube no. CH03). The tube was shaken immediately to allow mixture with lithium heparin and centrifuged 3 minutes. Total cholesterol was analyzed by a dry chemistry enzymatic method (Ektachem DT-60, Eastman Kodak Company, Rochester, New York). The machine was left stationary throughout the study. Trained laboratory technicians (n = 2) analyzed all samples and evaluated control samplesdaily to insure proper calibration. To assureaccuracy of finger stick measurements,venous samples were determined in approximately every tenth subject. Blood was first allowed to clot before centrifugation for 5 to 10 minutes. Serum was stored at between4Oto 6°C before transport to a Center for DiseaseControl standardizedcommercial referencelaboratory (Smith-Kline Laboratories). Dab tieis: Descriptive characteristics for the study sampleand pertinent subgroupswere determined. Linear regressionanalysis was usedto establish a correspondencebetween finger stick and venous cholesterol. Families were grouped according to the highest NCEP classof the parents: low = both parents had serum total cholesterol
It was hypothesized that healthy children wlth high cholesterol levels may have parents who exceed acceptabte cholesterol levels established by the National Chotesterol Education Program. One hundred slxty families (320 parents, 263 children aged 3 to 10 years) were evaluated for total cholesterol and other rlslc factors. Defore the study, almost half of the parents had net had serum total cholesterol measured. The odds ratlo for a child having a total chotesterol 25.17 mmoi/liter (266 mg/dl) was 13.&l (confidence interval 5.7 to 32.S) for a child with at teast 1 parent having cholesterol 26.26 mmol/Mter (246 mg/dl) versus a chitd whose parents had’low total cholesterol. Testing only chtldren who had at least 1 parent wlth a total chotesterol 25.17 mmol/liter (260 mg/dl) had a sensltlvity of 98% for detecting children’s total chotesterol1S.17 mmol/liter. It is concluded that parental total cholesterol is useful in ldentlfying children with high total cholesterol levels. Pediatricians may identify a large number of parents with hypercholesterolemia not previously recognized. (AmJCardid 1992;69:713-717)
hildren with elevated total cholesterol levels are likely to have elevated total cholesterol levels as adults.1-5In an effort to establish guidelines for identifying hypercholesterolemia in children, the National Cholesterol Education Program (NCEP) pediatric panel has suggested that children with a family history of heart disease,family history of hypercholesterolemia or parental serum total cholesterol 16.20 mmol/liter (240 mg/dl) be evaluatecL6Several studies have challenged this statement as being too restrictive, concluding that many children with hypercholesterolemia would be excluded from screening.7~8 Today, there is a lack of uniform strategy among pediatricians to identify children at risk for elevated cholesterol as adults?*iO To determine children at risk for elevatedcholesterol in a private pediatric office, we evaluated a variety of familial risk factors including parents’ serum total cholesterol to identify preadolescentchildren who should be evaluated for hypercholesterolemia. We hypothesized that children with high cholesterol would have parents who exceedacceptablecholesterol concentration established by the NCEP.” In addition, we hypothesized that comprehensivefamily screeningrather than assessment of individual children’s total cholesterol would be beneficial in identifying children and parents with hypercholesterolemia.
C
METHODS
Subjeetsr We asked 180 biologic families to enter the study, and completedata were obtained on 160 families with 263 children between the ages of 3 and 10 years. Data from these 160 families are included in this report. All were Caucasian.Children were identified at routine pediatric visits. Only parents and children who From the Children’s Memorial Hospital, Northwestern University were in good health were allowed entry. Pregnant womMedical School,Chicago, Illinois. This study was supportedin part by en or any subjectswith chronic illness, diabetes,renal or the Frieda Stone Heart Fund, Highland.Park, Illinois, and the Chilliver diseasewere excluded from participation because dren’s Memorial Institute for Researchand Education, Chicago, Illinois.Dr. Gidding is a memberof the FeinbergCardiovascularResearch of possible effects of these conditions on lipid levels.‘* Institute, Chicago, Illinois. ‘Manuscript received Septembci 3, 1991; Participants were recruited between June and Novemrevisedmanuscript receivedNovember 22,1991, and acceptedNovem- ber 1989 from a single group pediatric practice. ber 30. wre: A questionnaire was completed by Addressfor reprints: SamuelS. Gidding, MD, Division of Cardiology #21, Children’s Memorial Hospital, 2300 Children’s Plaza, Chica- both parents which asked about personal and family risk factors. Information was sought regarding prior go, Illinois 60614. PEDIATRIC CHOLESTEROL SCREENING
713
TABLE I Demographic Information on Participating Families Father No. of subjects
Age (yr) (range) Total cholesterol (mmol/liter) (range) Body mass index (kg/m*) (range) Family history of CAD (%) Family history of elevated total cholesterol (%) Tobacco use (%I No prior TC screen (%)
Mother
160 38 A 5 (27-61) 5.37 -c 1.8 (3.54-8.48) 25 r 3 (18-37) 48.8 36.6
160 36 r 4 (25-47) 5.01 k 0.9 (2.53-7.73) 23 f 4 (17-37) 51 33.1
11.9 43
Children 263 5.7 + 4.50 k 15.8 +
(149M/114F) 2.1 (3-10) 0.8 (2.71-8.40) 2 (10.4-25.3) 71 53
8.1 54
Values are mean + standard deviation. CAD = coronary artery disease; TC = total cholesterol. I
cholesterol testing of each parent, parental smoking, family history of coronary artery disease(angina, myocardial infarction, coronary bypass surgery or sudden cardiac death) and family history of elevatedtotal cholesterol or use of a lipid-lowering medication. A positive family history of coronary artery diseasewas defined as a positive responsefor that parent or a first-degree relative of either parent. A positive family history of elevated total cholesterol was defined as a report of total cholesterol 15.17 mmol/liter (200 mg/dl) in the parent or a first-degree relative of the parent. Clink~I aueumant: All subjectswere weighed and measuredusing office scales.Body mass index (weight [kg1/heightb211 was calculated for each subject.13 Nonfasting blood specimenswere obtained by finger stick. Freely flowing capillary blood was collected in a commercially available collection tube (Norfolk Scientific Collection Tube no. CH03). The tube was shaken immediately to allow mixture with lithium heparin and centrifuged 3 minutes. Total cholesterol was analyzed by a dry chemistry enzymatic method (Ektachem DT-60, Eastman Kodak Company, Rochester, New York). The machine was left stationary throughout the study. Trained laboratory technicians (n = 2) analyzed all samples and evaluated control samplesdaily to insure proper calibration. To assureaccuracy of finger stick measurements,venous samples were determined in approximately every tenth subject. Blood was first allowed to clot before centrifugation for 5 to 10 minutes. Serum was stored at between4Oto 6°C before transport to a Center for DiseaseControl standardizedcommercial referencelaboratory (Smith-Kline Laboratories). Dab tieis: Descriptive characteristics for the study sampleand pertinent subgroupswere determined. Linear regressionanalysis was usedto establish a correspondencebetween finger stick and venous cholesterol. Families were grouped according to the highest NCEP classof the parents: low = both parents had serum total cholesterol
