Investigation into childhood origins of coronary heart disease

1. Introduction

Post-mortem studies of children killed in road accidents have shown evidence of early degenerative changes in their arteries (Sanders, 1987). Post-mortem examination of Korean War casualties showed that 77% of U.S servicemen had some degree of atherosclerosis, with some 15% having considerable coronary narrowing (Enos, 1953). USA casualties in Vietnam showed 45% with some degree of atherosclerosis (McNamara, 1971).

The Coronary Prevention Group has stated that “the physiological conditions that predispose adults to CHD can occur during childhood: children can have raised blood pressure, raised cholesterol levels or be overweight.” and, say the American Health Foundation (1983), middle-aged atherosclerosis is related to the extent of fatty streaking in childhood, whilst other population studies have shown that levels of blood cholesterol in youth correlate with adult levels of cholesterol and incidence of CHD (American Health Foundation, 1983).

Image (339)

Figure 1.  Comparison of adult and childhood ‘ideal’ and ‘present’ means for total cholesterol.

It is accepted that atherosclerosis starts in childhood, is also associated with a diet high in saturated fat and low in polyunsaturates, and is therefore a paediatric problem (Widhalm, 1988; Sanders, 1987). Major changes occur in childhood for CHD risk with variations in BP, TC and HDL levels (Berenson, 1893; Lauer, 1988; Hofman, 1989; Mauer, 1991). Baseline HDL levels have been shown to be lowest in boys of poor physical fitness, declines in physical fitness associated with decreases in HDL levels (Hofman, 1989).

Children can be screened using recognised techniques with regard to major risk factors for later life CHD – lipids, blood pressure, obesity, level of physical activity. Also an important feature of disease suscept­ibility is the family history of CHD. In a survey of an adolescent population particular attention needs to be paid to familial aggregation of coronary risk factors, an aspect of CHD aetiology that has an impOrtant bearing on their future health.

2. Aims

There is still uncertainty surrounding the identification of hyperlipidaemia, hypertension, obesity in children yet by the age of ten many have already developed overt signs of predisposing factors of coronary heart disease (CHD). This underlines the necessity of more research into early screening to evaluate and define potential susceptib­ility to later coronary heart disease. The challenge becomes one of defining risk status more precisely (Lancet, 1991).

The proposed project would screen for major risk factors indicating possible susceptibility to CHD in 13 years plus school children. It is believed that community orientated prevention of CHD should start during childhood and that this requires established population distributions of lipid and blood pressure variables within young populations. In addition information gathered by this programme could be linked into diabetes research elswhere, with correlation with birthweights of the subjects. The study has both cross-sectional and longitudinal components – one in ten boys and one in ten girls will be selected for follow-up studies.

Firstly, the project will aim at establishing baseline data on demo­graphic, socioeconomic and geographical factors in relation to detected levels of primary biological risk factors for CHD (blood lipids, blood pressure etc), in a young premorbid population. The project would thus contribute to the development of “treatment strategies that are effect­ive and safe in young people.” (Lancet, 1991).

A second stage will provide an innovative evaluation of apoprotein AI (associated with ‘protective’ high density lipoprotein) and apoprotein B (associated with ‘atherogenic’ low density lipoprotein) to evaluate the value of these apolipoproteins as putative markers for adult risk of CHD. The dry chemistry reagent cartridges are still under development and will not be available for the first stage of the screening project.

3. Blood lipids in children

If serum lipids track over time then high risk profile adolescents would become high risk adults – in fact such tracking over time has been shown to be the case (Stein, 1981) with other longitudinal studies defining defining the tracking of lipid elevations into adult life (Lancet, 1991; Lauer, 1988). The Cincinnati Study (quoted in Stein, 1981) showed, on the on average, adolescents with high range cholesterol levels (e.g 4.65 mmol/1) are approximating adult levels (e.g. 6.2 mmol/1). An Ohio Study showed 20% of 3-18 year olds above total cholesterol 4.8 mmol/1 (Garcia, 1989).

The Heartbeat Wales Clinical Survey of 1985 showed children aged 12-19 years had mean cholesterol levels of 4.42 mmol/1 in boys and 4.67 in girls (Heartbeat Wales, 1985). In a Northern Ireland survey some 20-25% of young adults aged 18-21 years had blood cholesterol levels above the WHO recommended optimum of 5.2 mmol/1 (Brown, 1989). For people under 20 years every cholesterol measure­ment higher than 5.2 mmol/1 (200mg/d1) must be considered elevated (Widhalm, 1988).

Numerous epidemiological surveys of children indicate that “differences in cholesterol and triglyceride levels appear early in childhood among cultures, among race-sex groups within cultures, and among socioeconomic groups within cultures.” (Stein, 1981). Figure 1 compares adult and child­hood ‘ideal’ and present means for total cholesterol according to the WHO (1990).

It has been estimated that children who are heterozygous for the inherited condition of familial hypercholesterolaemia (FH) have a mean serum cholesterol level of 7.75 mmol/1 or 300mg/d1 (Kwiterowich, 1974). FH confers susceptibility to premature coronary heart disease and underlies about 4% of heart attacks occurring during middle-age (Coronary Prevention Group, 1988). The condition can be identified in childhood and has a frequency of about 1:500 individuals. Elevated lipids can also be found in less common inherited hyperlipidaemic conditions, including familial combined hyperlipidaemia (FCH).

Evidence shows that it is worthwhile not only to screen children (as well as a necessity in terms of preventive cardiology) but also to track them (if possible) over a period of years – especially those found with lipid levels above 5.17 mmol/1 if not a possible optimum level of 4.65 mmol/1.

4. Blood pressure in children

There is considerable variability in childhood blood pressure with blood pressure in adolescents being far more labile than adults, nonetheless ‘normal’ values have been established in relation to sex and age in children (Purcell, 1985). In the UK 1 in 100 children have mild persistent essential hypertension, with 1 in 500 with evidence of severe secondary hypertension (Chantler, 1983). Hypertension in children is regarded as factual when a persistent BP of 130/90 mm Hg is present – by such criteria 1 or 2% of children are hypertensive (Catzel, 1984).

Studies have also shown that tracking of children with elevated BP’s still show high levels in later years (Purcell, 1985). Chantler suggests (1985) that “all British children should have their blood pressure measured by the school nurse at five and eleven years.” ence.” (Stein, 1981).

Height and body mass are significant determinants of blood pressure and obesity (Purcell, 1985),It is believed that blood pressure patterns “may therefore be established from an early age, which would be consistent with reports that emphasise the importance of childhood in the origins of cardiovascular disease.” (Whincup, 1988). There is some tendency for children with raised blood pressure to become hypertensive as adults.” (Coronary Prevention Group, 1988).

5. Methods

There are difficulties in the assessment of risk factors for CHD in in children for physiological and anatomic reasons – risk variable parameters vary with age due to early life hormonal and biochemical processes that are connected with growth and development (Brown, 1989). For this reason it is therefore important to establish a large body of baseline data via a population screening survey. In addition, the lipid testing would be of a non-fasting free-living population – though to obtain greater accuracy all blood tests would be obtained prior to lunch time.

Surveys of school pupils provide a comprehensive profile of the distrib­ution of the major risk factors across an adolescent population as well as help illuminate class, ethnic and regional variations and indicate areas for action in disease prevention and health education. Secondly surveys would collect data on the geographic and social origin of individuals, their parents, and possibly grandparents. This would enable factors such as spatial and social mobility to be analysed in association with the prevalence of risk variables.

The project would comprise 2 parts: clinical evaluation and a questionaire, directed at 13+ years pupils in three of the six Upper Schools of Oxford City. The clinical assessment would be of a minimally invasive nature that would involve (1) anthropometric measurements; (2) BP recording, (3) two finger stick blood samples. The questionaire would seek to establish information on the subject’s (1) demographic, (2) socioeconomic and (3) geographical back­ground in order to account for social class, origin and family history of CHD. An initial draft of the questionaire is shown as Appendix 1. The following biological variables will be sought:

  1. Blood total cholesterol (TC)
  2. Triglycerides (TG)
  3. High density lipoprotein (HDL)
  4. Low density lipoprotein (LDL) by estimation
  5. Blood glucose
  6. LDL/HDL Ratio by calculation
  7. TC/LDL Ratio by calculation
  8. Mean systolic blood pressure
  9. Mean diastolic blood pressure
  10. Mean pulse rate
  11. Blood group (ABO)
  12. Weight, height, skin-fold thicknesses, waist, hips and right arm circumference
  13. Body mass index calculation

Anthropometric records would be done with stadiometer, metric tapes, digital scales, and Harpenden type skinfold calipers. Blood pressure would also be recorded by a digital electronic sphygmomanometer (DES) using a COPAL UA 751, taking the mean of two readings with appropriate size cuffs.

The blood sample measurements will be done using the portable Abbott Diagnostics ‘Vision’ autoanalyser (offered for 18 months use) for lipids and glucose (using four cartridges in the Abbot ‘Cardiac Evaluation’ pack). Portable autoanlyzer use removes the need to rely on expensive and time consuming laboratory analysis, is minimally invasive and ethically of an acceptable nature especially where children are concerned. Blood samples would be obtained using an Autolet II capillary blood sampler using Unilet stylets with platforms (Autolet ‘white’ for young skin) appropriate for thumb skin puncture. See Appendix 1 for a preliminary cost analysis of the project.

6. Analysis

Analysis of baseline data would use the SPSS-X statistics package using appropriate analyses of variance (ANOVA), multivariate analyses of variance (MANOVA), significance tests, establishment of confidence intervals and the and calculation of odds ratios and relative risk (‘R’) as required.

7. Sample

The six Upper Schools in Oxford City will comprise a forecast population of 3,490 13-18 year-olds in 1992 (the year of the proposed screening survey) of which 2187 will be 13-18 in the three selected upper schools – Cherwell, Cheney, and Peers Schools – which would provide an even geographical spread across the City. The sample of 2187 comprises some 62% of the upper school population within the City. The sample size is determined by costs and the feasibility of screening all upper school children within three school terms. Forecast pupil populations are shown in Table 1 with school location within the City boundary illustrated in Figure 2.

Image (340)

Table 1.  Oxford City Upper School Forecasts 1990-93.

Image (341)

Figure 2.  Map of Oxford showing location of City’s Upper Schools (13-18 years).

Image (342)

Appendices 1 to 4.  Survey forms for the investigation.

Image (343)

Image (345)

Image (344)


American Health Foundation, 1983, Summary and recommendations of the conference on blood lipids in children; optimal levels for early prevention pf coronary heart disease. Prev. Med 12, 728-40

Berenson, G.S, 1983, Clinical and anatomic correlates with cardio­vascular disease in children from the Bogalusa Heart Study, in Schettler, F. et al. (eds), Atherosclerosis IV, Berlin, 1983.

Brown, J.S, 1989, Screening for coronary heart disease: the sooner the better? Modern Medicine, April 1989, p279.

Catzel P & G Roberts, 1984, A Short Textbook of Paediatrics, 2nd ed, p147, Hodder & Stoughton.

Chantler C, 1983, Paediatric hypertension. Midwife, Health Visitor and Community Nurse. Feb 19

Coronary Prevention Group, 1988, Children at risk: should prevention of coronary heart disease begin in childhood? A Policy Statement from the Scientific and Medical Advisory Committee.

Enos W.F. et al, 1953, Coronary disease among United States soldiers killed in action in Korea: preliminary report. JAMA, 152, 1090-4.

Garcia R.E & D.S Moodie, 1989, Routine cholesterol surveillance in childhood. Pediatrics, 84, 751-55.

Heartbeat Wales, 1985, Heartbeat Report No 20, Heart of Wales: Clinical Results of the Welsh Heart Health Survey, 1985, Directorate of the Welsh Heart Health Programme, Cardiff.

Hofman, A, 1989. The association between physical fitness and cardio­vascular disease in children in a five year follow-up study. Int.J.Epidem. 18, (4), 831-35, 1989.

Kwiterowich P.O et al, 1974, Familial hypercholesterolaemia (one form of familial type II hyperlipoproteinaemia): a study of its biochemical, genetic, and clinical presentation in childhood. J. Clin. Invest. 53, 1237-1249.

Lancet, 1991, Cholesterol screening in childhood, Editorial, 29.6.91.

Lauer R.M, et al, 1988, Factors affecting the relationship between childhood and adult cholesterol levels: the Muscatine Study. Pediatrics, 82, 309-18.

McNamara J.J, et al, 1971, Coronary artery disease in combat casualties in Vietnam. JAMA, 216, 1185-87.

Mauer, 1991. Should there be intervention to alter serum lipids in children? Ann.Rev.Nutr. 11, 375-91, 1991.

Purcell H, 1985, Blood pressure and hypertension in children. Card. in Practice, January, 8-9.

Sanders T.A.B, 1987, What should our children eat? J.Roy.Soc.Health, 107, 6, Dec, 222-23.

Stein E.A, et al, 1981, Coronary risk factors in the young. Ann.Rev.Med. 32, 601-13.

Whincup P.H, et al, 1988, Blood pressure in British children: associations with adult blood pressure and cardiovascular maortality, Lancet, Oct 15, 890-93.

WHO, 1990, Prevention in childhood and youth of adult cardiovascular diseases: Time for Action. WHO, Geneva, 792.

Widhalm K, 1988, Dietary treatment of hypercholesterolaemia in children: recent aspects., in Widhalm & Naito (eds).


The proposal (1993) was accepted by the Department of Biology at Oxford Brookes University, as well as laboratory facilities, and supervision. A qualified practice nurse to carry out tests was also nominated. The cooperation of the Lipid Clinic of the John Radcliffe Hospital, Oxford was agreed. Abbot Laboratories agreed to supply the portable analyser and technical backup, training plus supplies of lipid analysis modules. The response from Oxford City Council Department of environmental Health was somewhat muted despite the existence of the Oxford Heart Health Project.

Eventually the project was unable to raise the necessary finance and studentships from medical research organisations and councils to inaugurate and carry out the research. Local authorities, both health and educational, were also muted in their responses, stating that childhood diabetes and obesity were not a problem in Oxford or likely to be in the future.


Leave a comment

Filed under Research

Discussion & Comment Welcome

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s