Acta PrYdiatr Scand Suppl373: 66-72, 1991
Cholesterol: Should W e Screen All Children or Change the Diet of All Children JUNE K. LLOYD From the Institute of Child Health, University of London, UK
ABSTRACT. Lloyd, J. K. (Institute of Child Health, University of London, UK). Cholesterol: Should we screen all children or change the diet of all children. Acta Paediatr Scand Suppl373: 66, 1991. Hypercholesterolaemia is a major risk factor for coronary heart disease and may present during childhood. Dietary measures can reduce plasma cholesterol and may thus delay or prevent the development of the atherosclerotic process. Although plasma cholesterol concentrations measured during childhood track into adult life with a correlation coefficient of about 0.6 this in itself is insufficient to justify total population screening of children especially as the mechanisms for management and follow-up and their social, psychological and economic implications have not been adequately evaluated. Targeted screening of children in families with the genetic disorder of familial hypercholesterolaemia, where the risk of premature coronary heart disease is very high, should, however, be undertaken even though such screening may only identify half of all affected children. Dietary change designed to lower plasma cholesterol can be applied to the whole population including children over the age of 2 years, does not require pre-determination of plasma cholesterol, and is to be recommended. The effects of such change on the growth and health of children should be monitored.
INTRODUCTION Coronary heart disease with its underlying pathology of atherosclerosis and thrombosis remains a major cause of mortality and morbidity in all so called developed societies, and is responsible for one third of all British deaths in middle life (1). Its prevention is a matter that demands attention at an earlier age. In this respect the possibility of screening for risk factors during childhood has been advocated; this article will address the problem of cholesterol screening. As long ago as 1953 Enos and his colleagues highlighted the frequency and severity of coronary lesions in young, outwardly healthy American soldiers killed in combat in Korea (2). Nearly two decades later MacNamara reported a similar situation in American soldiers killed in Vietnam (3) and more recently Stary has shown early coronary lesions in 17 O/o of American children dying before the age of 5 (4). In the Bogalusa Heart Study aortic fatty streaks in young people dying before the age of 25 years (mean at death 18 years) were shown to be strongly correlated with ante-mortem levels of both total and low-density lipoprotein cholesterol ( 5 ) . Such evidence is not confined to studies from the United States; a WHO Report in 1982 shows prevalence rates of atherosclerotic plaques at age 20 varying between 10 and 30 O/o in a number of different populations with only Lima and Mexico at 5 O/o or lower (6). Accepting that the pathological process often started in childhood-and without entering into the controversy of whether or not the presence of fatty streaks was meaningful, or whether they were reversible, or what the nature of the ultimate occlusive process was-Kannel & Dawber in 1972 proposed a five point plan for paediatricians to follow in order to play their part in the prevention of atherosclerosis (7). Their recommendations were based on the known risk factors for coronary
Acta Paediatr Scand Suppl 373
Cholesterol, screening and diet
heart disease which had been identified in epidemiological studies in adults: hypercholesterolaemia, hypertension, cigarette smoking, physical inactivity and obesity. Although the present discussion is concerned only with the detection and treatment of hyperlipidaemia (and even more specifically with hypercholesterolaemia) it cannot be over emphasised that atherosclerosis and coronary heart disease have a multifactorial aetiology and that any approach at prevention must recognise this. Nearly 20 years have passed since Kannel & Dawber made their proposals and many similar exhortations have been made in the intervening period. One of the latest is contained in a further report from a WHO Expert Committee entitled “Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action” (8). In relation to the control of hypercholesterolaemia in children various courses of action are possible. Amongst these are the screening of all children for blood cholesterol concentration followed by “treatment” of those with hypercholesterolaemia, and a change in the diet of all children which would be expected to result in lowering of blood cholesterol of all children. These approaches are of course not necessarily mutually exclusive but for ease of presentation will be considered separately. POPULATION SCREENING FOR CHOLESTEROL IN CHILDHOOD Before embarking on the screening of all children for plasma cholesterol concentrations assurance is needed that the degree of tracking is sufficient, that an appropriate cut-off point can be chosen, and that management and follow up arrangements for those children identified as at risk are in place. Assurance is also needed that the technical, administrative, and financial aspects have been thoroughly worked out; a detailed discussion of the latter aspects are beyond the scope of this article.
Tracking The phenomenon of tracking of plasma cholesterol concentrations through childhood and into adult life has been the subject of a number of studies. In Australia, Boulton and colleagues have recently reported their results of an investigation of tracking and of parent child associations of total plasma cholesterol, low density lipoprotein (LDL) cholesterol and high density lipoprotein (HDL) cholesterol from infancy through to the age of 13 years (9). The degree of tracking from 3 months to 13 years showed a steady increase in the coefficient of correlation over the years to an r value of 0.76 between 8 and 13 years. An investigation of the continuity through time for children in the upper and lower quintiles for total cholesterol at the age of 2 years, showed that a greater proportion of those in the uppermost quintile remained there (47 O/o) than those in the lower quintile who remained in their track (26O/o). The Muscatine Study has now reported a 15-year follow-up of school children tested at 6-1 5 years (10). For those children tested at the age of 7 to 8 years and followed up to ages 20-25 years, the r value for total cholesterol was 0.56 for males and 0.64 for females (similar figures were obtained for LDL cholesterol but there was no significant correlation with HDL). For those children tested between the ages of 13-14 years and followed to between 26-30 years the r value for both males and females was 0.52 for total cholesterol and not significantly different for LDL cholesterol. Of those children with cholesterol in childhood greater than the 90th centile, 43 O/o had levels above the 90th centile between ages 20-30. From this study the authors were able to calculate that between 25-5Oo/o of the variability of total and LDL cholesterol levels in adults could be explained by childhood levels.
67
68
J. K . Lloyd
Acta Paediatr Scand Suppl 373
Whether the degree of tracking shown in these and other studies is sufficient to warrant testing of the whole childhood population is a matter of opinion; it would seem clear, however, that estimations would have to be repeated-at the very least in those children whose levels were in the upper centiles-at intervals of perhaps 3 or 4 years, and that more secure information would be provided by estimations made in later childhood. Cut-off levels
If total population screening were to be adopted, a decision would have to be made as to the appropriate cut-off level above which children would be selected for management and follow-up. To simply take the upper decile or quintile of the population under study could be inappropriate if mean cholesterol concentrations in the population being screened were already unacceptably high. To use a population distribution around the proposed ideal mean of 2.85 mmol/l (with a standard deviation of 0.52 mmol/l) (6) would entail the vast majority of children in many populations being singled out for special attention. Even accepting a feasible mean of 3.62 mmol/I (with a standard deviation of 0.65 mmol/l) (6) would also result-at least in the USA and Britain-in a large number of children being selected if a figure of 2 standard deviations above this mean (4.92 mmol/l) were adopted. For practical reasons a decision to use the upper quintile in populations known to have high mean levels would probably be used, and cut-off values would have to be kept under regular review.
Management and follow-up Children identified by a single estimation of total plasma concentration above an arbitrary cut-off point at whatever age would initially require a repeat blood sample, taken in the fasting state and by venepuncture (screening samples could almost certainly rely on a skin prick specimen) and the estimation of not only total cholesterol but also LDL and HDL cholesterol concentrations. The children emerging from this second examination with raised LDL cholesterol (again an appropriate cut-off point would have to be determined) would then enter a dietary treatment regime designed to lower their plasma cholesterol concentrations to a “satisfactory” level-also needing to be defined. The efficacy of the diet prescribed and compliance with the regime, would require monitoring by regular repeat estimations of, at the very least, total cholesterol. No information is available on which to decide the appropriate interval between repeat estimations, or regarding the degree of compliance that can be expected from children and their families identified and managed in this manner. A regime as outlined, even if follow-up estimations were only undertaken once a year (and this would almost certainly be inadequate, at least in the early stages) would have major implications for health professionals including dietitians and laboratory staff. The psychological impact of being labelled ‘At Risk‘ for adult heart disease on both the chiId concerned and the family has not been quantified and cannot be ignored. The argument that knowing the cholesterol improves compliance with a management regime that sets a child apart from a peer group needs to be validated. Conclusions In my opinion the degree of tracking of plasma cholesterol is by itself insufficient to warrant a population screening programme for all children. Further studies may
Acta Paediatr Scand Suppl 373
Chotesterol, screening and diet
indicate the best age at which the initial estimation should be made and define the most appropriate cut-off point. A much clearer understanding is required of the implications of the consequent management regime which has resource implications for the health services of the country, and psychological implications for the children and their families. The need for research in these and other areas is clearly highlighted in the latest WHO report (8). SCREENING FOR FAMILIAL HYPERCHOLESTEROLAEMIA Children who are heterozygous for the dominantly inherited disorder of familial hypercholesterolaemia are at special risk for the early development of atherosclerosis and subsequent coronary heart disease (1 1). Cholesterol levels in such children track extremely well and treatment of affected children is now generally accepted as desirable, even though direct proof that such treatment will delay the atheromatous process is lacking. Nevertheless, the accumulated evidence from intervention studies in adults together with data derived from children homozygous for the disorder (12) support the desirability of treatment. Compliance with both dietary and current drug therapies is, however, not satisfactory ( 1 3 ) and much work remains to be done regarding the best treatment regimes for these children. In spite of these uncertainties most authorities now agree that every opportunity should be taken to detect the disorder. Because there is no precise cut-off point for plasma cholesterol or LDL above which the diagnosis can be made with certainty, population screening cannot be advocated as a means of detecting familial hypercholesterolaemia. A targeted screening approach through families is thus the current method suggested for detecting asymptomatic children. Even this approach is not without its difficulties. Some children will have borderline results and the diagnosis remain in doubt. It has also been estimated that as many as 50% of affected children in a population will not be identified through a family history approach (14). In spite of this, screening targeted at high risk families remains an appropriate measure and all children in such families should be examined. DIETARY CHANGE FOR ALL CHILDREN Many studies both in infants and older children have shown that plasma cholesterol concentrations can be lowered by changes in their diet. In early infancy children fed on human milk have the highest plasma cholesterol concentrations whereas those receiving a formula in which the fat is largely of vegetable origin and rich in polyunsaturated fatty acids, and the total cholesterol concentration is low, have the lowest concentrations (15). There is, however, no evidence that the early feeding regime affects cholesterol concentrations in later childhood or adult life (1 6-19). The overwhelming advantages of breast feeding (20) and optimal nutrition and growth (21) during the early months of life have led to general agreement that no modifications of current feeding practices are warranted under the age of 2 years (22). For children over the age of 2 years, evidence is lacking that a change in diet that would lower the total fat content to about 30-35 O/o of total calories from the current levels of over 40 %, with an increase in the consumption of polyunsaturated fat and a reduction in total cholesterol and salt intake, would have any deleterious effects. Indeed such a diet is normal for children in some countries. Evidence is, however,
69
70
J. K . Lloyd
Acta Paediatr Scand Suppl 373
also lacking that such a diet would of itself result in slowing of the atherosclerotic process and a reduction in adult coronary heart disease. The evidence suggesting that these desirable effects would occur is based on epideiniological studies such as the 16 nation study reported by Knuiinan ( 2 3 ) , studies in adults (24, 25) and in experimental animals ( 2 6 ) . Given the lack of clear evidence it is not surprising that advice varies. The American Academy of Pediatrics in 1986 concluded that “any recommendations for changing toward a more restrictive dietary pattern during the first 2 decades of life should await demonstration that such dietary restrictions are needed and, in addition, that such restrictions would support adequate growth and development for children and adolescents” (27). The American Heart Association ( 2 8 ) , The American Health Foundation (29), a Consensus Development Panel sponsored by the National Institutes of Health in the USA (22), the Department of Health in the UK (30), and the Coronary Prevention Group in the UK all recommend a dietary policy for children (over age 2 ) directed towards achieving lower plasma cholesterol concentrations. The diets of free living, healthy children cannot be taken in isolation from the diet of the whole population in which they are growing up. If the diet of the adult population is to change in the ways currently suggested, and there is some evidence that this is likely to happen although perhaps not to the degree recommended, then the diets of children should change similarly. This will require firm policy decisions, as indicated by the Coronary Prevention Group, and will have implications for professionals advising families, manufacturers and their marketing practices, and for those who provide food for children outside their homes. Although it is generally assumed that children acquire their food preferences in their homes and that such preferences are relatively stable, there is little evidence for this assertion. The resemblance between parents and children in relation to the acquisition of stable food preferences is actually very low (31) and although early experiences of food may affect later food choices there is no evidence that early experiences are more important than later ones ( 3 2 ) . Furthermore there is evidence that food preferences of children are strongly influenced by food preferences of their peer group and those whom they greatly admire, and that these effects are long lasting ( 3 3 ) . This is a further reason for suggesting that if it is desirable to lower the plasma cholesterol levels of some children by dietary change, the changes are most likely to succeed if the measures are directed towards all children. The recent report that socioeconomic conditions in childhood are significantly associated with ischaemic heart disease during middle age (34) highlights the importance of ensuring that dietary advice (and other risk factor avoidance advice) reaches this disadvantaged section of the child community, even if there is no hard evidence as to the reason(s) for the association. It also emphasises the importance of monitoring the effects of any recommendations to ensure that optimum growth is maintained and that no harm comes to any child. REFERENCES 1. Mortality statistics. OPCS Series D H 5 No 12 London: HMSO, 1985. 2. Enos WF, Holmes RH, Beyer JC. Coronary disease among United States soldiers killed in action in Korea. JAMA 1953; 152: 1090-93. 3. McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coronary artery disease in combat casualties in Vietnam. JAMA 1971; 216: 1185-87. 4. Stary HC. Macrophages, macrophage foam cells, and eccentric intimal thickening in the coronary arteries of young children. Atherosclerosis 1987; 64: 91-108.
Acta Paediatr Scand Suppl 373
Cholesterol, screening and diet
5. Newman WP, Freedman DS, Voors AW et al. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J Med 1986; 314: 138-44. 6. WHO Technical Report Series. Prevention of Coronary Heart Disease. N o 678. 1982. 7. Kannel WB, Dawber TR. Atherosclerosis as a pediatric problem. J Pediatr 1972; 80: 544-54. 8. WHO Technical Report Series. Prevention in childhood and youth of adult cardiovascular diseases: time for action. N o 792. 1990. 9. Boulton TJC, Magarey A, Cockington R. Cholesterol from infancy to age thirteen: tracking and parent-child associations. In: Annales Nest16 1989; 48: 70-76. 10. Lauer RM, Lee J, Clarke WR. Factors affecting the relationship between childhood and adult cholesterol levels: the Muscatine Study. Pediatrics 1988; 82: 309-1 8. 1 1. Slack J. Risks of ischaemic heart disease in familial hyperlipoproteinaemic states. Lancet 1969; ii: 1380-82. 12. West RJ, Gibson PJ, Lloyd JK. Treatment of homozygous familial hypercholesterolaemia: an informative sibship. Br Med J 1985; 2: 1079-80. 13. West RJ, Lloyd JK, Leonard JV. Long-term follow-up of children with familial hypercholesterolaemia treated with cholestyramine. Lancet 1980; ii: 873-75. 14. Stare TJ, Belamarich PF, Shea S, Dobrin BE, Gersony WM, Deckelbaum RJ. Family history fails to identify many severely hypercholesterolemic children. Circulation 1989; Suppl 11, 80: 262. 15. Andersen GE, Lifschitz C, Friis-Hansen B. Dietary habits and serum lipids during the first 4 years of life. A study of 95 Danish children. Acta Paediatr Scand 1979; 68: 165-70. 16. Huttunen JK, Saarinen UM, Kostiainen E, Siimes MA. Fat composition of infant diet does not influence subsequent serum lipid levels in man. Atherosclerosis 1983; 46: 87-94. 17. Glueck CJ, Tsang R, Balistireri W, Fallat R. Plasma and dietary cholesterol in infancy: effects of early low or moderate dietary cholesterol intake on subsequent response to increased dietary cholesterol. Metabolism 1972; 21 : 1 181-92. 18. Hodgson PA, Ellefson RD, Elveback LR, Harris LE, Nelson RA, Weidman WH. Comparison of serum cholesterol in children fed high, moderate, or low cholesterol milk diets during the neonatal period. Metabolism 1976; 25: 739-46. 19. Marmot MG, Page CM, Atkins E, Douglas JW. Effect of breast-feeding on plasma cholesterol and weight in young adults. J Epidemiol Community Health 1980; 34: 164-67. 20. Department of Health and Social Security. Present day practice in infant feeding: third report. Report on Health and Social Subjects, N o 32. London: Her Majesty’s Stationery Office, 1988. 21. Barker DJ, Winter PD, Osmond C, Margetts B, Simmonds SJ. Weight in infancy and death from ischaemic heart disease. Lancet 1989; ii: 577-80. 22. Consensus Development Panel: lowering blood cholesterol to prevent heart disease. JAMA 1985; 253: 2080-86. 23. Knuiman JT, Hermus RJJ, Hautvast JGAJ. Serum total and high density lipoprotein (HDL) concentrations in rural and urban boys from 16 countries. Atherosclerosis 1980; 36: 529-37. 24. Tuomilehto J, Geboers J, Salonen JT, Nissinen A, Kuulasmaak K, Puska P. Decline in cardiovascular mortality in North Karelia and other parts of Finland. Br Med J 1986; 293: 1068-7 1. 25. Sytkowski PA, Kannel WB, Dagostin RB. Changes in risk factors and the decline in mortality from cardiovascular disease: the Framingham Heart Study. N Engl J Med 1990; 322: 1635-41. 26. Mott GE, Jackson EM, McMahan CA, Farley CM, McGill HC Jr. Cholesterol metabolism in juvenile baboons: influence of infant and juvenile diets. Arteriosclerosis 1985; 5 : 347-54. 27. American Academy of Pediatrics, Committee on Nutrition. Prudent life style for children: dietary life style and cholesterol. Pediatrics 1986; 78: 521-25. 28. Weidman W, Kwiterovich P Jr, Jesse MJ et al. Diet in the healthy child: AHA Nutrition Committee Report. Circulation 1983; 67: 141 1A-1414A. 29. Wynder EL, Berenson GS, Epstein F H et al. Summary and recommendations of the conference on blood lipids in children: optimal levels for early prevention of coronary artery disease. Prev Med 1983; 12: 728-40. 30. Department of Health and Social Security. Committee on Medical Aspects of Food
71
72
31. 32. 33. 34.
J. K. Lloyd
Acta Paediatr Scand Suppl 373
Policy. Diet and cardiovascular disease. Report on Health and Social Subjects. No. 28. London: Her Majesty’s Stationary Office, 1984. Rozin P. Acquisition of stable food preferences. Nutrition Reviews 1990; 48: 106-13. Beauchamp GK. Ontogenesis of taste preference. In: Walker D, Kretchmer N, eds. Foods, nutrition and evolution. New York: Masson, 1981: 49-57. Birch LL. The acquisition of food acceptance patterns in children. In: Boakes RA, Popperwell DA, Burton MJ eds. Eating habits: food physiology and learned behaviour. Chichester: Wiley, 1986: 107-1 30. Kaplan GA, Salonen JT. Socioeconomic conditions in childhood and ischaemic heart disease during middle age. Br Med J 1990; 301: 1121-23.
(J. K. L.) Institute of Child Health 30 Guilford Street London W C l N 1EH UK
Cholesterol: Should W e Screen All Children or Change the Diet of All Children JUNE K. LLOYD From the Institute of Child Health, University of London, UK
ABSTRACT. Lloyd, J. K. (Institute of Child Health, University of London, UK). Cholesterol: Should we screen all children or change the diet of all children. Acta Paediatr Scand Suppl373: 66, 1991. Hypercholesterolaemia is a major risk factor for coronary heart disease and may present during childhood. Dietary measures can reduce plasma cholesterol and may thus delay or prevent the development of the atherosclerotic process. Although plasma cholesterol concentrations measured during childhood track into adult life with a correlation coefficient of about 0.6 this in itself is insufficient to justify total population screening of children especially as the mechanisms for management and follow-up and their social, psychological and economic implications have not been adequately evaluated. Targeted screening of children in families with the genetic disorder of familial hypercholesterolaemia, where the risk of premature coronary heart disease is very high, should, however, be undertaken even though such screening may only identify half of all affected children. Dietary change designed to lower plasma cholesterol can be applied to the whole population including children over the age of 2 years, does not require pre-determination of plasma cholesterol, and is to be recommended. The effects of such change on the growth and health of children should be monitored.
INTRODUCTION Coronary heart disease with its underlying pathology of atherosclerosis and thrombosis remains a major cause of mortality and morbidity in all so called developed societies, and is responsible for one third of all British deaths in middle life (1). Its prevention is a matter that demands attention at an earlier age. In this respect the possibility of screening for risk factors during childhood has been advocated; this article will address the problem of cholesterol screening. As long ago as 1953 Enos and his colleagues highlighted the frequency and severity of coronary lesions in young, outwardly healthy American soldiers killed in combat in Korea (2). Nearly two decades later MacNamara reported a similar situation in American soldiers killed in Vietnam (3) and more recently Stary has shown early coronary lesions in 17 O/o of American children dying before the age of 5 (4). In the Bogalusa Heart Study aortic fatty streaks in young people dying before the age of 25 years (mean at death 18 years) were shown to be strongly correlated with ante-mortem levels of both total and low-density lipoprotein cholesterol ( 5 ) . Such evidence is not confined to studies from the United States; a WHO Report in 1982 shows prevalence rates of atherosclerotic plaques at age 20 varying between 10 and 30 O/o in a number of different populations with only Lima and Mexico at 5 O/o or lower (6). Accepting that the pathological process often started in childhood-and without entering into the controversy of whether or not the presence of fatty streaks was meaningful, or whether they were reversible, or what the nature of the ultimate occlusive process was-Kannel & Dawber in 1972 proposed a five point plan for paediatricians to follow in order to play their part in the prevention of atherosclerosis (7). Their recommendations were based on the known risk factors for coronary
Acta Paediatr Scand Suppl 373
Cholesterol, screening and diet
heart disease which had been identified in epidemiological studies in adults: hypercholesterolaemia, hypertension, cigarette smoking, physical inactivity and obesity. Although the present discussion is concerned only with the detection and treatment of hyperlipidaemia (and even more specifically with hypercholesterolaemia) it cannot be over emphasised that atherosclerosis and coronary heart disease have a multifactorial aetiology and that any approach at prevention must recognise this. Nearly 20 years have passed since Kannel & Dawber made their proposals and many similar exhortations have been made in the intervening period. One of the latest is contained in a further report from a WHO Expert Committee entitled “Prevention in Childhood and Youth of Adult Cardiovascular Diseases: Time for Action” (8). In relation to the control of hypercholesterolaemia in children various courses of action are possible. Amongst these are the screening of all children for blood cholesterol concentration followed by “treatment” of those with hypercholesterolaemia, and a change in the diet of all children which would be expected to result in lowering of blood cholesterol of all children. These approaches are of course not necessarily mutually exclusive but for ease of presentation will be considered separately. POPULATION SCREENING FOR CHOLESTEROL IN CHILDHOOD Before embarking on the screening of all children for plasma cholesterol concentrations assurance is needed that the degree of tracking is sufficient, that an appropriate cut-off point can be chosen, and that management and follow up arrangements for those children identified as at risk are in place. Assurance is also needed that the technical, administrative, and financial aspects have been thoroughly worked out; a detailed discussion of the latter aspects are beyond the scope of this article.
Tracking The phenomenon of tracking of plasma cholesterol concentrations through childhood and into adult life has been the subject of a number of studies. In Australia, Boulton and colleagues have recently reported their results of an investigation of tracking and of parent child associations of total plasma cholesterol, low density lipoprotein (LDL) cholesterol and high density lipoprotein (HDL) cholesterol from infancy through to the age of 13 years (9). The degree of tracking from 3 months to 13 years showed a steady increase in the coefficient of correlation over the years to an r value of 0.76 between 8 and 13 years. An investigation of the continuity through time for children in the upper and lower quintiles for total cholesterol at the age of 2 years, showed that a greater proportion of those in the uppermost quintile remained there (47 O/o) than those in the lower quintile who remained in their track (26O/o). The Muscatine Study has now reported a 15-year follow-up of school children tested at 6-1 5 years (10). For those children tested at the age of 7 to 8 years and followed up to ages 20-25 years, the r value for total cholesterol was 0.56 for males and 0.64 for females (similar figures were obtained for LDL cholesterol but there was no significant correlation with HDL). For those children tested between the ages of 13-14 years and followed to between 26-30 years the r value for both males and females was 0.52 for total cholesterol and not significantly different for LDL cholesterol. Of those children with cholesterol in childhood greater than the 90th centile, 43 O/o had levels above the 90th centile between ages 20-30. From this study the authors were able to calculate that between 25-5Oo/o of the variability of total and LDL cholesterol levels in adults could be explained by childhood levels.
67
68
J. K . Lloyd
Acta Paediatr Scand Suppl 373
Whether the degree of tracking shown in these and other studies is sufficient to warrant testing of the whole childhood population is a matter of opinion; it would seem clear, however, that estimations would have to be repeated-at the very least in those children whose levels were in the upper centiles-at intervals of perhaps 3 or 4 years, and that more secure information would be provided by estimations made in later childhood. Cut-off levels
If total population screening were to be adopted, a decision would have to be made as to the appropriate cut-off level above which children would be selected for management and follow-up. To simply take the upper decile or quintile of the population under study could be inappropriate if mean cholesterol concentrations in the population being screened were already unacceptably high. To use a population distribution around the proposed ideal mean of 2.85 mmol/l (with a standard deviation of 0.52 mmol/l) (6) would entail the vast majority of children in many populations being singled out for special attention. Even accepting a feasible mean of 3.62 mmol/I (with a standard deviation of 0.65 mmol/l) (6) would also result-at least in the USA and Britain-in a large number of children being selected if a figure of 2 standard deviations above this mean (4.92 mmol/l) were adopted. For practical reasons a decision to use the upper quintile in populations known to have high mean levels would probably be used, and cut-off values would have to be kept under regular review.
Management and follow-up Children identified by a single estimation of total plasma concentration above an arbitrary cut-off point at whatever age would initially require a repeat blood sample, taken in the fasting state and by venepuncture (screening samples could almost certainly rely on a skin prick specimen) and the estimation of not only total cholesterol but also LDL and HDL cholesterol concentrations. The children emerging from this second examination with raised LDL cholesterol (again an appropriate cut-off point would have to be determined) would then enter a dietary treatment regime designed to lower their plasma cholesterol concentrations to a “satisfactory” level-also needing to be defined. The efficacy of the diet prescribed and compliance with the regime, would require monitoring by regular repeat estimations of, at the very least, total cholesterol. No information is available on which to decide the appropriate interval between repeat estimations, or regarding the degree of compliance that can be expected from children and their families identified and managed in this manner. A regime as outlined, even if follow-up estimations were only undertaken once a year (and this would almost certainly be inadequate, at least in the early stages) would have major implications for health professionals including dietitians and laboratory staff. The psychological impact of being labelled ‘At Risk‘ for adult heart disease on both the chiId concerned and the family has not been quantified and cannot be ignored. The argument that knowing the cholesterol improves compliance with a management regime that sets a child apart from a peer group needs to be validated. Conclusions In my opinion the degree of tracking of plasma cholesterol is by itself insufficient to warrant a population screening programme for all children. Further studies may
Acta Paediatr Scand Suppl 373
Chotesterol, screening and diet
indicate the best age at which the initial estimation should be made and define the most appropriate cut-off point. A much clearer understanding is required of the implications of the consequent management regime which has resource implications for the health services of the country, and psychological implications for the children and their families. The need for research in these and other areas is clearly highlighted in the latest WHO report (8). SCREENING FOR FAMILIAL HYPERCHOLESTEROLAEMIA Children who are heterozygous for the dominantly inherited disorder of familial hypercholesterolaemia are at special risk for the early development of atherosclerosis and subsequent coronary heart disease (1 1). Cholesterol levels in such children track extremely well and treatment of affected children is now generally accepted as desirable, even though direct proof that such treatment will delay the atheromatous process is lacking. Nevertheless, the accumulated evidence from intervention studies in adults together with data derived from children homozygous for the disorder (12) support the desirability of treatment. Compliance with both dietary and current drug therapies is, however, not satisfactory ( 1 3 ) and much work remains to be done regarding the best treatment regimes for these children. In spite of these uncertainties most authorities now agree that every opportunity should be taken to detect the disorder. Because there is no precise cut-off point for plasma cholesterol or LDL above which the diagnosis can be made with certainty, population screening cannot be advocated as a means of detecting familial hypercholesterolaemia. A targeted screening approach through families is thus the current method suggested for detecting asymptomatic children. Even this approach is not without its difficulties. Some children will have borderline results and the diagnosis remain in doubt. It has also been estimated that as many as 50% of affected children in a population will not be identified through a family history approach (14). In spite of this, screening targeted at high risk families remains an appropriate measure and all children in such families should be examined. DIETARY CHANGE FOR ALL CHILDREN Many studies both in infants and older children have shown that plasma cholesterol concentrations can be lowered by changes in their diet. In early infancy children fed on human milk have the highest plasma cholesterol concentrations whereas those receiving a formula in which the fat is largely of vegetable origin and rich in polyunsaturated fatty acids, and the total cholesterol concentration is low, have the lowest concentrations (15). There is, however, no evidence that the early feeding regime affects cholesterol concentrations in later childhood or adult life (1 6-19). The overwhelming advantages of breast feeding (20) and optimal nutrition and growth (21) during the early months of life have led to general agreement that no modifications of current feeding practices are warranted under the age of 2 years (22). For children over the age of 2 years, evidence is lacking that a change in diet that would lower the total fat content to about 30-35 O/o of total calories from the current levels of over 40 %, with an increase in the consumption of polyunsaturated fat and a reduction in total cholesterol and salt intake, would have any deleterious effects. Indeed such a diet is normal for children in some countries. Evidence is, however,
69
70
J. K . Lloyd
Acta Paediatr Scand Suppl 373
also lacking that such a diet would of itself result in slowing of the atherosclerotic process and a reduction in adult coronary heart disease. The evidence suggesting that these desirable effects would occur is based on epideiniological studies such as the 16 nation study reported by Knuiinan ( 2 3 ) , studies in adults (24, 25) and in experimental animals ( 2 6 ) . Given the lack of clear evidence it is not surprising that advice varies. The American Academy of Pediatrics in 1986 concluded that “any recommendations for changing toward a more restrictive dietary pattern during the first 2 decades of life should await demonstration that such dietary restrictions are needed and, in addition, that such restrictions would support adequate growth and development for children and adolescents” (27). The American Heart Association ( 2 8 ) , The American Health Foundation (29), a Consensus Development Panel sponsored by the National Institutes of Health in the USA (22), the Department of Health in the UK (30), and the Coronary Prevention Group in the UK all recommend a dietary policy for children (over age 2 ) directed towards achieving lower plasma cholesterol concentrations. The diets of free living, healthy children cannot be taken in isolation from the diet of the whole population in which they are growing up. If the diet of the adult population is to change in the ways currently suggested, and there is some evidence that this is likely to happen although perhaps not to the degree recommended, then the diets of children should change similarly. This will require firm policy decisions, as indicated by the Coronary Prevention Group, and will have implications for professionals advising families, manufacturers and their marketing practices, and for those who provide food for children outside their homes. Although it is generally assumed that children acquire their food preferences in their homes and that such preferences are relatively stable, there is little evidence for this assertion. The resemblance between parents and children in relation to the acquisition of stable food preferences is actually very low (31) and although early experiences of food may affect later food choices there is no evidence that early experiences are more important than later ones ( 3 2 ) . Furthermore there is evidence that food preferences of children are strongly influenced by food preferences of their peer group and those whom they greatly admire, and that these effects are long lasting ( 3 3 ) . This is a further reason for suggesting that if it is desirable to lower the plasma cholesterol levels of some children by dietary change, the changes are most likely to succeed if the measures are directed towards all children. The recent report that socioeconomic conditions in childhood are significantly associated with ischaemic heart disease during middle age (34) highlights the importance of ensuring that dietary advice (and other risk factor avoidance advice) reaches this disadvantaged section of the child community, even if there is no hard evidence as to the reason(s) for the association. It also emphasises the importance of monitoring the effects of any recommendations to ensure that optimum growth is maintained and that no harm comes to any child. REFERENCES 1. Mortality statistics. OPCS Series D H 5 No 12 London: HMSO, 1985. 2. Enos WF, Holmes RH, Beyer JC. Coronary disease among United States soldiers killed in action in Korea. JAMA 1953; 152: 1090-93. 3. McNamara JJ, Molot MA, Stremple JF, Cutting RT. Coronary artery disease in combat casualties in Vietnam. JAMA 1971; 216: 1185-87. 4. Stary HC. Macrophages, macrophage foam cells, and eccentric intimal thickening in the coronary arteries of young children. Atherosclerosis 1987; 64: 91-108.
Acta Paediatr Scand Suppl 373
Cholesterol, screening and diet
5. Newman WP, Freedman DS, Voors AW et al. Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J Med 1986; 314: 138-44. 6. WHO Technical Report Series. Prevention of Coronary Heart Disease. N o 678. 1982. 7. Kannel WB, Dawber TR. Atherosclerosis as a pediatric problem. J Pediatr 1972; 80: 544-54. 8. WHO Technical Report Series. Prevention in childhood and youth of adult cardiovascular diseases: time for action. N o 792. 1990. 9. Boulton TJC, Magarey A, Cockington R. Cholesterol from infancy to age thirteen: tracking and parent-child associations. In: Annales Nest16 1989; 48: 70-76. 10. Lauer RM, Lee J, Clarke WR. Factors affecting the relationship between childhood and adult cholesterol levels: the Muscatine Study. Pediatrics 1988; 82: 309-1 8. 1 1. Slack J. Risks of ischaemic heart disease in familial hyperlipoproteinaemic states. Lancet 1969; ii: 1380-82. 12. West RJ, Gibson PJ, Lloyd JK. Treatment of homozygous familial hypercholesterolaemia: an informative sibship. Br Med J 1985; 2: 1079-80. 13. West RJ, Lloyd JK, Leonard JV. Long-term follow-up of children with familial hypercholesterolaemia treated with cholestyramine. Lancet 1980; ii: 873-75. 14. Stare TJ, Belamarich PF, Shea S, Dobrin BE, Gersony WM, Deckelbaum RJ. Family history fails to identify many severely hypercholesterolemic children. Circulation 1989; Suppl 11, 80: 262. 15. Andersen GE, Lifschitz C, Friis-Hansen B. Dietary habits and serum lipids during the first 4 years of life. A study of 95 Danish children. Acta Paediatr Scand 1979; 68: 165-70. 16. Huttunen JK, Saarinen UM, Kostiainen E, Siimes MA. Fat composition of infant diet does not influence subsequent serum lipid levels in man. Atherosclerosis 1983; 46: 87-94. 17. Glueck CJ, Tsang R, Balistireri W, Fallat R. Plasma and dietary cholesterol in infancy: effects of early low or moderate dietary cholesterol intake on subsequent response to increased dietary cholesterol. Metabolism 1972; 21 : 1 181-92. 18. Hodgson PA, Ellefson RD, Elveback LR, Harris LE, Nelson RA, Weidman WH. Comparison of serum cholesterol in children fed high, moderate, or low cholesterol milk diets during the neonatal period. Metabolism 1976; 25: 739-46. 19. Marmot MG, Page CM, Atkins E, Douglas JW. Effect of breast-feeding on plasma cholesterol and weight in young adults. J Epidemiol Community Health 1980; 34: 164-67. 20. Department of Health and Social Security. Present day practice in infant feeding: third report. Report on Health and Social Subjects, N o 32. London: Her Majesty’s Stationery Office, 1988. 21. Barker DJ, Winter PD, Osmond C, Margetts B, Simmonds SJ. Weight in infancy and death from ischaemic heart disease. Lancet 1989; ii: 577-80. 22. Consensus Development Panel: lowering blood cholesterol to prevent heart disease. JAMA 1985; 253: 2080-86. 23. Knuiman JT, Hermus RJJ, Hautvast JGAJ. Serum total and high density lipoprotein (HDL) concentrations in rural and urban boys from 16 countries. Atherosclerosis 1980; 36: 529-37. 24. Tuomilehto J, Geboers J, Salonen JT, Nissinen A, Kuulasmaak K, Puska P. Decline in cardiovascular mortality in North Karelia and other parts of Finland. Br Med J 1986; 293: 1068-7 1. 25. Sytkowski PA, Kannel WB, Dagostin RB. Changes in risk factors and the decline in mortality from cardiovascular disease: the Framingham Heart Study. N Engl J Med 1990; 322: 1635-41. 26. Mott GE, Jackson EM, McMahan CA, Farley CM, McGill HC Jr. Cholesterol metabolism in juvenile baboons: influence of infant and juvenile diets. Arteriosclerosis 1985; 5 : 347-54. 27. American Academy of Pediatrics, Committee on Nutrition. Prudent life style for children: dietary life style and cholesterol. Pediatrics 1986; 78: 521-25. 28. Weidman W, Kwiterovich P Jr, Jesse MJ et al. Diet in the healthy child: AHA Nutrition Committee Report. Circulation 1983; 67: 141 1A-1414A. 29. Wynder EL, Berenson GS, Epstein F H et al. Summary and recommendations of the conference on blood lipids in children: optimal levels for early prevention of coronary artery disease. Prev Med 1983; 12: 728-40. 30. Department of Health and Social Security. Committee on Medical Aspects of Food
71
72
31. 32. 33. 34.
J. K. Lloyd
Acta Paediatr Scand Suppl 373
Policy. Diet and cardiovascular disease. Report on Health and Social Subjects. No. 28. London: Her Majesty’s Stationary Office, 1984. Rozin P. Acquisition of stable food preferences. Nutrition Reviews 1990; 48: 106-13. Beauchamp GK. Ontogenesis of taste preference. In: Walker D, Kretchmer N, eds. Foods, nutrition and evolution. New York: Masson, 1981: 49-57. Birch LL. The acquisition of food acceptance patterns in children. In: Boakes RA, Popperwell DA, Burton MJ eds. Eating habits: food physiology and learned behaviour. Chichester: Wiley, 1986: 107-1 30. Kaplan GA, Salonen JT. Socioeconomic conditions in childhood and ischaemic heart disease during middle age. Br Med J 1990; 301: 1121-23.
(J. K. L.) Institute of Child Health 30 Guilford Street London W C l N 1EH UK
