Family Practice Vol. 16, No. 5, 501-505
© Oxford University Press 1999
The advisability of implementing cholesterol screening in school-age children and adolescents with a family history of cardiovascular disease and hyperlipidaemia
a Department of Family Medicine, China Medical College Hospital,
b Department of Clinical Laboratory, China Medical College Hospital and
c Department of Medical Research, China Medical College Hospital, Taichung, Taiwan.
Dr Lin Cheng-Chieh, Department of Family Medicine, China Medical College Hospital, 2, Yuh-Der Road, Taichung, Taiwan.
Liu C-S, Lin C-C, Shih H-C and Li T-C. The advisability of implementing cholesterol screening in school-age children and adolescents with a family history of cardiovascular disease and hyperlipidaemia.Family Practice 1999; 16:501–505.
Received 25 September 1998; Revised 9 April 1999; Accepted 13 May 1999.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. The family basis of coronary heart disease is well recognized and it is important for family physicians to assess whether children have elevated cholesterol levels.
Objectives. We aimed to evaluate the advisability of implementing cholesterol screening in children with a family history of cardiovascular disease and hyperlipidaemia.
Methods. We conducted a cross-sectional study in Taiwan from February to June 1996. There were 47 800 students in the population. A total of 4520 students were recruited by two-stage sampling. All the participants were required to fill out a structured questionnaire.
Results. The response rate was 92.5%. Our results show that 16–18% of the children had a positive family history of cardiovascular disease or hyperlipidaemia. Children with a family history of hyperlipidaemia were significantly more likely to have elevated total cholesterol and low-density lipoprotein cholesterol than those without such a history (both odds ratios: 1.4, P < 0.05). Positive predictive values of hyperlipidaemia were less than 13% based on family history. More than 75% of children with abnormal lipid levels would be missed.
Conclusions. The data suggest that parents' self-reported family history is an ineffective means of identifying children with elevated serum lipid levels in Taiwan. Further research and modification of current National Cholesterol Education Program Panel guidelines for selective cholesterol screening in children may be warranted.
Keywords. Children, cholesterol, family history, screening.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Coronary heart disease (CHD) is the leading cause of death in Taiwan1 and other countries.2 There is little doubt that elevated blood cholesterol, especially elevated low-density lipoprotein cholesterol, is an important risk factor for CHD.3 The atherosclerotic process begins early in life, decades before patients may become symptomatic.4 Previous research has demonstrated that cholesterol levels track well from childhood to adulthood.5 This association enables early identification of hypercholesterolaemia and warrants appropriate intervention in children and adolescents.
The family basis of CHD is well recognized,3 and it is important for family physicians to be able to assess whether children or adolescents in a family with risk of CHD have elevated cholesterol levels for their age. Strategies for intervention concerning the assessment of lipid levels in children vary in different organizations. In the USA, the National Cholesterol Education Program (NCEP) has set a standard for community and individual intervention.6 The NCEP Children's Panel recommends that children and adolescents over 2 years of age with a family history of cardiovascular disease (CHD, peripheral vascular disease, cerebrovascular disease, premature sudden cardiac death) or hypercholesterolaemia (total cholesterol 
6.21 mmol/l) should be tested for dyslipidaemia.
However, the effectiveness of such screening strategies is still controversial.7,8 The purpose of this study is to evaluate the suitability of implementing cholesterol screening in children and adolescents in Taiwan with a family history of CHD, cerebrovascular disease (CVA) and hyperlipidaemia, as suggested by the NCEP.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We conducted a cross-sectional study in Taichung, Taiwan from February to June 1996. The population consisted of first and fourth-graders of primary schools and first-graders of junior high schools. There were a total of 47 800 students in the population during the time of the study. A two-stage sampling design was used to draw subjects, with equal probability for each sampling unit within each stage. In the first stage of sampling, six schools were randomly selected from a total of 79. In the second stage, all students in the selected schools were included in our study, by cluster sampling. A total of 4520 subjects were selected in our study, of whom 4183 agreed to participate. Thus the overall response rate was 92.5%.
All selected students were given a structured questionnaire with an invitation letter briefly describing the study and instructions on filling out the questionnaire to take home to their parents. Those parents who agreed to participate in this study would sign their consent form before they filled out questionnaires. Information collected by questionnaire consisted of demographic factors and medical history of premature (<55 years of age) CHD and CVA, and history of high cholesterol among the first-degree relatives of the students' parents and grandparents. The questions of CHD and CVA asked for the information diagnosed by a physician. The prevalence of peripheral vascular disease is pretty low, so we do not have enough statistical power to estimate its effect. Therefore, the questionnaire did not ask about the information of peripheral vascular disease.
The students fasted for more than 8 hours on the day of examination. The blood samples were sent to the clinical laboratory of China Medical College Hospital within 4 hours for analysis. Total cholesterol (TC) and triglyceride (TG) were determined by the enzymatic method using commercial kits (Boehringer Inc., USA). High-density lipoprotein cholesterol (HDL
C) was measured by the direct method9 using commercial kits (Kyowa Co., Japan). All the analyses were performed by Hitachi-735 autoanalyzer. Low-density lipoprotein cholesterol (LDLC) was calculated using Friedewald's formula: LDLC = TC – (HDLC + TG/5).
Hypercholesterolaemia was defined as a TC level greater than 5.18 mmol/l (200 mg/dl), as suggested by NCEP. Hypertriglyceridaemia was defined as a TG level greater than 1.70 mmol/l (150 mg/dl). Abnormal HDLC and LDLC values were defined as an HDLC level lower than 0.90 mmol/l (35 mg/dl), and an LDLC level greater than 3.35 mmol/l (130 mg/dl), respectively.
Serum variables approached a normal distribution (Kolmogorov–Smirnov test). Comparisons among groups were based on the analysis of variance (ANOVA) for independent variables. Student's t-tests were used to compare the differences of continuous data. The sensitivity, specificity, positive predictive value and odds ratio were calculated using standard methodology from 2 x 2 table analysis. The statistical analyses were performed with the aid of an SAS package (Version 6.06, SAS Institute Inc., Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
52.9% of the subjects were boys. The mean age was 10.1 years. The mean TC level was 4.23 mmol/l (163.2 mg/dl); 9.6% of the students had a TC level greater than 5.18 mmol/l. The means of TG, LDL
C and HDLC were 0.86 mmol/l (76.3 mg/dl), 2.28 mmol/L (88.2 mg/dl) and 1.55 mmol/l (59.8 mg/dl), respectively.
Table 1
displays the percentage of children with positive histories for cardiovascular disease and hyperlipidaemia in various family members. A positive parental history was rare for CHD and CVA (2.5 and 0.5%, respectively), and a history of parental hyperlipidaemia was low (6.2%). A positive grandparental history was much more common: CHD and CVA had occurred in at least one grandparent for 14.7 and 15.2%, respectively, of the participants, and a grandparental hyperlipidaemia was found in 14.0%. However, approximately 27% of the participants responded that they were uncertain of their family history.
|
Serum levels of TC, TG, LDL
C and HDLC of children with different family histories are shown in Table 2. Children with a positive family history of hyperlipidaemia had significantly higher levels of TC and LDLC than those without such a history. The means of TC and LDLC for those with and without a family history of hyperlipidaemia were 4.36 mmol/l (168.2 mg/dl) versus 4.22 mmol/l (162.8 mg/dl) and 2.39 mmol/l (92.1 mg/dl) versus 2.28 mmol/l (88.0 mg/dl), respectively (both P < 0.01). Children with a positive family history of CHD and CVA did not show significantly higher serum lipid levels than those without such family histories.
|
Children with a family history of hyperlipidaemia were significantly more likely to have elevated levels of serum total cholesterol and LDL
C than those without such a history (both odds ratios: 1.4, P < 0.05) (Table 3). A family history of vascular diseases (CHD, CVA) was not associated with an increased risk for abnormal serum levels of TC, TG, LDLC or HDLC. Table 3 also shows that the values of sensitivity for serum lipid screening were all less than 20%, based on family history of vascular diseases (CHD, CVA). A slightly higher sensitivity was noted based on family history of hyperlipidaemia (21.8–23.8%). The values of specificity ranged from 81.5 to 84.6%. Positive predictive values were quite low based on family history (<12.5%).
|
With the variables family history, sex, age and obesity of the children as determinant factors, the odds ratios were calculated for each abnormal serum lipid level as the dependent variables using logistic regression models (Table 4
). Older children were significantly less likely than younger children to have abnormal serum TC, LDLC and HDLC levels (odds ratios 0.8–0.9, all P < 0.01). Obese children were significantly more likely to have abnormal serum TC, TG and LDLC levels than non-obese children (odds ratios 1.6, 6.6 and 2.0, respectively, all P < 0.01). Children with a positive family history of hyperlipidaemia were also more likely to have abnormal serum TC levels than those without such a history (odds ratio 1.5, P < 0.05). Factors such as sex, family history of CVA and family history of CHD were not associated with an increased risk of abnormal serum lipid levels (odds ratio 0–1.0).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The American National Cholesterol Education Program (NCEP) Children's Panel6 has advocated screening for high blood cholesterol in children and adolescents, but only among those at high risk as judged by a positive family history. The main finding of this study is that in Taiwan, targeted screening based on family history is likely to be highly insensitive and inaccurate. Children with a positive family history of hyperlipidaemia were more likely to have elevated TC and LDL
C levels than those without such a history. Family history of coronary heart disease and cerebrovascular disease were not significantly associated with abnormal serum lipid levels.
The prevalence of TC 5.18 mmol/l was 9.6% in this study. The sensitivity of serum lipid screening was 11.6–23.8% for various target conditions. Previously, the sensitivity of total cholesterol screening based on family history was between 10 and 60% in different studies.7,8 The positive predictive values of the family history were quite low, with positive predictive values ranging from 2.6 to 12.5%. The low positive predictive values may be explained, in part, by low sensitivity and low prevalence of the target condition.
A family with a history of cardiovascular disease or hyperlipidaemia might remember the history more clearly than those without such histories. Comparisons of lipid values between those who reported a family history and those who did not, reveal no significant difference (not presented in the paper), which reduces, but does not rule out, the possibility of bias. If systematic misclassification resulting from the ‘don't know’ category having a higher proportion of family history would result in overestimation of its effect, then systematic misclassification resulting from the ‘don't know’ category having a lower proportion of family history would result in underestimation of its effect. In our case, the potential bias may come from the latter, which may be responsible for the weak associations observed in our study. There are two other alternative explanations for such a weak association. One explanation is that parents' self-reported histories were inaccurate.10 In the present study, family history was obtained from a self-reporting questionnaire completed in the home. It is unclear whether the questionnaire was completed by both of the parents. A second possible explanation is that parents who had a positive family history may have been more likely to be following a prudent diet, and their children may have already begun to lower their cholesterol levels. This may have contributed to the low yield of family history as a screening guideline. When considering family history as a guideline for screening cholesterol levels in children, the risk of having abnormal serum lipid levels would be underestimated.
Self-reported family history factors appear to be an ineffective means of identifying children with elevated serum lipids levels.7,8 Using the current NCEP guidelines that only children whose parents have a positive history of hypercholesterolaemia, premature CHD and CVA should be screened, we found that more than 75% of children with abnormal serum lipid levels would be missed.
The NCEP Pediatric Panel have projected that their selective screening guidelines will identify 41% of children with elevated LDLC, whereas Benuck et al.11 found that clinical hypercholesterolaemia (not by health history alone) in either parent identified 52% of children with TC values greater than 5.18 mmol/l. The authors suggest that family physicians measure parents' blood lipid or obtain records of cardiovascular disease history before determining children's family risk status. However, both of these interventions are somewhat impractical in the general population. Inclusive population screening would be the most effective approach if thorough identification of children with elevated serum lipid levels is desired.12 Use of a self-reported family history of cardiovascular disease and hyperlipidaemia should be discouraged.
|
Acknowledgments |
|---|
This study was financially supported by the China Medical College Hospital, Taiwan. The authors acknowledge the expert assistance of Dr TT Chen, Dr SW Lai, Ms IM Fu, Ms CF Chen, Ms WF Chen, Ms SR Chuan and Ms HH Tien.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Department of Health, Executive Yuan, R.O.C. Health and Vital Statistics 1995 (1). General Health Statistics. Taipei: Department of Health, 1996.
2 Health United States 1989. Hyattsville, MD: US Dept of Health and Human Services, 1989; 21: 104. DHHS publication 90-1232.
3 Moll PP, Sing CF, Weidman WH, Gordon H, Ellefson RD, Hodgson PA. Total cholesterol and lipoproteins in school children. Prediction of coronary heart disease in adult relatives. Circulation 1983; 67: 127–134.
4
McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coronary artery disease in combat casualties in Vietnam. JAMA 1971; 216: 1185–1187.
5 Boulton TJC, Magarey AM, Cockington RA. Tracking of serum lipids and dietary energy, fat and calcium intake from 1 to 15 years. Acta Paediatr 1995; 84: 1050–1055.[Web of Science][Medline]
6
National Cholesterol Education Program. Report of the expert panel on blood cholesterol levels in children and adolescents. Pediatrics 1992; 89: 525–584.
7
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: 347–353.
8
Steiner NJ, Neinstein LS, Pennbridge J. Hypercholesterolemia in adolescents: Effectiveness of screening strategies based on selected risk factors. Pediatrics 1991; 88: 269–275.
9
Sugiuchi H, Miyauchi K. Direct measurement of high density lipoprotein cholesterol in serum with polyethylene glycol-modified enzymes and sulfated 
-cyclodextrin. Clin Chem 1995; 41: 713–723.
10 Hastrup JL, Hotchkiss AP, Johnson CA. Accuracy of knowledge of family history of cardiovascular disorders. Health Psychol 1985; 4: 291–306.[Web of Science][Medline]
11 Benuck I, Gidding SS, Donovan M et al. Usefulness of parental serum total cholesterol levels in identifying children with hypercholesterolemia. Am J Cardiol 1992; 69: 713–717.[Web of Science][Medline]
12 Bistrtzer T, Batash D, Barr J et al. Routine childhood screening for hyperlipidemia in Israel. Isr J Med Sci 1996; 32: 725–729.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||