Prospective Randomized Trial of Low-Saturated-Fat, Low-Cholesterol Diet During the First 3 Years of Life
The STRIP Baby Project
Background The long-term consequences of modified fat intake in early childhood are poorly known. The randomized prospective STRIP baby project evaluates the effects of repeated dietary counseling on nutrient intakes and serum lipid values in children 7 months to 3 years old.
Methods and Results One thousand sixty-two infants were randomized to intervention and control groups at 7 months of age. The families of the 540 intervention children were counseled to reduce the child's intake of saturated fat and cholesterol but to ensure adequate energy intake. Five hundred twenty-two control children consumed an unrestricted diet. Food records were kept, and serum lipids were measured at 5- to 12-month intervals. Intakes of saturated fat, fat as proportion of energy (E%), and cholesterol were lower in the intervention children than in control children at 13, 24, and 36 months of age. Fat intake by the intervention children decreased from 29±5 E% at 8 months of age to 26±6 E% at 13 months and then increased to 30±5 E% at 24 months and to 31±5 E% at 36 months. The control children consumed 29±4 E%, 28±5 E%, 33±5 E%, and 33±5 E% of fat at 8, 13, 24, and 36 months, respectively. The ratio of dietary polyunsaturated to saturated fats of the intervention children was consistently higher than that of the control children (P<.0001). Baseline adjusted mean serum cholesterol concentration was lower in the intervention children than control children between 13 and 36 months (P<.0001; 95% confidence interval of the difference between the group means, −0.27 to −0.12 mmol/L). The effect was significant only in boys (95% confidence interval, −0.39 to −0.20 mmol/L in boys; −0.21 to 0.01 mmol/L in girls).
Conclusions Repeated individualized dietary counseling markedly reduces the increase in serum cholesterol concentration that occurs in control children during the first years of life.
Reduction of serum cholesterol concentration by means of dietary modification is probably the most important population strategy for prevention of CHD. Because the genetic and biochemical mechanisms involved in the development of vascular atherosclerotic lesions are active immediately after birth, it might be feasible to introduce a low-saturated-fat, low-cholesterol diet in early childhood. Such an approach is further supported by findings suggesting that the extent of preatherosclerotic lesions in coronary arteries and aorta depends on serum lipoprotein concentrations even in children and young adults. In the Bogalusa Heart Study, antemortem total and LDL cholesterol concentrations predicted the extent of postmortem aortic fatty streaks, and VLDL cholesterol concentrations correlated closely with coronary fatty streaks in 7- to 24-year-old subjects.1 Postmortem lipid concentrations and arterial atherosclerotic lesions were also tightly linked in 15- to 34-year-old males in the PDAY study.2 According to a meta-analysis of studies in adults, the benefit of serum cholesterol reduction relates to the age at which the reduced serum cholesterol concentration is achieved.3 Furthermore, a study of schoolboys in 16 countries showed high serum cholesterol concentrations in countries with high CHD incidence in adults.4
Dietary counseling in childhood may have a crucial effect on CHD risk later in life through permanent modification of eating habits and preferences. Despite frequently expressed fears that modification of fat intake in infants and young children might induce growth failure5 and the current recommendations that a low-fat, low-cholesterol diet is unsuitable for children <2 years of age,6 7 we recently showed that serum cholesterol concentrations were markedly lower in infants fed a low-saturated-fat, low-cholesterol diet compared with values in children consuming a free diet, and growth proceeded at a normal rate in infancy.8 To evaluate how effectively dietary changes introduced in infancy persist during early childhood when families receive repeated counseling, we extended the randomized prospective STRIP baby trial to the age of 3 years. We now present the results of the dietary surveys and serum lipids, which suggest that continuous dietary counseling at an early age is both feasible and effective.
Initial recruitment of families into the STRIP baby trial was performed by nurses at the well-baby clinics in the city of Turku, Finland (160 000 inhabitants). At the infants' routine 5-month visits, all families were informed about the possibility of taking part in the study. One thousand fifty-four families with 1062 infants (56.5% of the eligible age cohort) were voluntarily enrolled. When the infant was 6 months old, a pediatrician met the family for the first time and explained the design of the trial. After the parents had given informed consent, the infants were allocated by random numbers at 7-month visits between March 1990 and May 1992 to intervention (n=540; 284 boys) and control (n=522; 266 boys) groups. The intervention children visited the counseling team at the ages of 7, 8, 10, 13, 15, 18, 21, 24, 30, and 36 months and the control children at the ages of 7, 13, 18, 24, 30, and 36 months. Nonfasting blood samples were drawn at the ages of 7, 13, 24, and 36 months for measurement of serum cholesterol, HDL cholesterol, and apo A-I and B concentrations as well as of dry-chemistry cholesterol. Three-day or 4-day food records of the child's diet were kept at 8, 13, 24, and 36 months.
The children continued regular visits at the city's well-baby clinics for vaccinations, follow-up of growth and development, and health education. Infants in both groups were advised to receive daily supplementation of 400 μg (1000 IU) vitamin A and 10 μg (400 IU) vitamin D2 until the age of 2 years. Infants were breast-fed for 5±4 months (mean±SD) in both groups, and solid foods were introduced between the ages of 3 and 5 months.
The trial was approved by the Joint Committee on Ethics of Turku University and Turku University Central Hospital. Informed consent was obtained from the families.
Five pediatricians, three dietitians, and a registered nurse took part in counseling of the families. The pediatrician recorded the child's past history, examined the child, and evaluated the development of the child by questioning and clinical examination. The dietitian examined the child's dietary history, explained how the child's food consumption could be accurately recorded, and suggested changes to the composition and amounts of foods consumed by the intervention children.
Counseling of the Intervention Group
Optimal diet for the child was defined to contain energy according to the child's hunger, protein 10 to 15 E%, carbohydrates 50 to 60 E%, and fat 30 to 35 E%, with a P/M/S ratio of 1:1:1. The intervention mothers were encouraged to continue breast-feeding as long as they found it feasible. Breast milk or formula was suggested to be continued until the age of 12 months, and later, 0.5 to 0.6 L of skim milk was suggested for daily use. The mothers were taught to replace the “missing” milk fat in the child's daily diet with vegetable oil (usually 2 to 3 teaspoons; mainly low–erucic acid rapeseed oil containing 6% saturated, 57% monounsaturated, and 32% polyunsaturated fats9 ) or soft margarine. The aim was to maintain fat intake at the same level as in children who used milk with 1.9% fat. Use of oil or soft margarine in food preparation was encouraged. Detailed suggestions were made to replace products containing large amounts of saturated fat with products containing fat of better quality. Use of leaner meat products, low-fat cheese, and ice cream with vegetable fat and ample use of vegetables were advised. No foods were prohibited, but daily use of candy or foods containing large amounts of simple sugars was discouraged. Fish was recommended as one of the main meals once or twice a week. A fixed diet was never ordered, but suggestions of small changes leading the diet toward the optimal composition were made during each visit.
Counseling of the Control Group
Parents of the control children received the basic health education routinely given at Finnish well-baby clinics. On the basis of current recommendations, cow's milk with at least 1.9% fat was suggested for use after the age of 1 year. No detailed suggestions concerning dietary fat were given, and dietary issues were discussed only briefly.
At the ages of 8 and 13 months, when the daily variations in the child's diet were small, the parents, usually mothers (and if pertinent, personnel in the day-care center), recorded the child's food consumption for 3 consecutive days. At 24 and 36 months of age, when the child's diet had become more diversified, the recording was extended to 4 consecutive days. On each occasion, the record included at least 1 weekend day. Attempts to reschedule the dates of food recording were made if the child had an acute illness.
After randomization of the children to the intervention and control groups at the age of 7 months, a dietitian instructed the parents how to record the child's food intake. The parents first recorded the child's food intake at the age of 8 months. At that time, almost 60% of the children received formula as their only milk source, in addition to solid foods. Because we had no means of accurately estimating breast-milk consumption or composition in such a large number of infants, analysis of true nutrient intakes was possible only in formula-fed children at the age of 8 months and in all subjects from the age of 13 months on, when all children had been weaned.
Before the child was 8 months old, dietary advice had been minimal in the intervention families as well. Nutrient intakes at 8 months of age showed no significant differences between the intervention and control groups. Consequently, we used these 8-month nutrient intakes as baseline intake values for the children. Food intake was recorded next at 13 months of age, when the children had changed from breast milk or formula to cow's milk.
In counseling, the dietitians used written instructions with drawings of portions of some ordinary foods. Household measures (eg, spoons, cups, glasses) were also used, or, alternatively, the foods were weighed with a kitchen scale. In practice, most parents and members of the day-care personnel used household measures. The food records were evaluated on each visit by the dietitians. Brand names and food preparation methods were taken to increase the accuracy of recording. The time interval between the recording days and the evaluation was <1 week. An experienced dietitian transferred the data into a computer. Nutrient compositions were analyzed with the Micro-Nutrica PC program,10 which uses the Food and Nutrient Data Base of the Social Insurance Institution.9 This program calculates 62 nutrients of more than 650 of the most commonly used foods in Finland, including baby foods and formulas.
Blood was drawn from the antecubital vein by use of minimal stasis. Cutaneous analgesia was attained with lidocaine and prilocaine cream (EMLA, Astra). If the first puncture attempt was unsuccessful, only a finger-prick sample was taken for dry-chemistry serum cholesterol analysis. After clotting of the venous blood sample at room temperature for 30 to 60 minutes and low-speed centrifugation (at 3400g for 12 minutes), serum was separated and stored at −25°C for a period of <1 month. Freezing as such has no effect on cholesterol concentrations, and normolipemic serum can be stored at −20°C for at least 11 weeks before analysis.11
Serum cholesterol concentration was measured by a fully enzymatic method12 (CHOD-PAP, Merck) in an AU 510 automatic analyzer (Olympus) except for the samples analyzed before November 1990, which were measured with a Boehringer CHOD-PAP kit13 and an OLLI analyzer (Kone). Serum HDL cholesterol concentration was measured after precipitation of LDL and VLDL with dextran sulfate 500 000.14 The results for samples analyzed before November 1990 were corrected according to the follow-ing equations: corrected cholesterol=1.068859×cholesterol(1990)−0.317158 and corrected HDL cholesterol=1.066248×HDL cholesterol (1990)−0.030667. Serum apos A-I and B were determined immunoturbidimetrically with Apo A-I and B kits (Orion Diagnostica).15 The measurements were standardized against WHO International Reference Materials SP1-01 for apo A-I and SP3-07 for apo B. Results for samples analyzed before September 1993 were corrected according to the following equations: corrected apo A-I=1.022726×apo A-I(during 1990 through 1993)+0.047643 and corrected apo B=1.056093×apo B(during 1990 through 1993)+0.041142. The measurements were done at the laboratory of the Research and Development Unit of the Social Insurance Institution, Turku, Finland. The laboratory continuously cross-checks lipid determinations with the WHO reference laboratory in Prague.
The interassay (intra-assay) coefficients of variation of the total cholesterol, HDL cholesterol, and apo A-I and B determinations were 2.0% (1.5%), 1.9% (1.2%), 3.0% (1.8%), and 4.5% (3.3%), respectively.
Dry-chemistry cholesterol concentrations (Reflotron, Boehringer Mannheim) were measured in all venous blood samples and used immediately in counseling of the intervention families. The dates of blood drawings were rather evenly distributed throughout the year, and samples were collected at the time of the day most convenient to the family, usually between 9 am and 6 pm. Because of the young age of the subjects, the samples were nonfasting. Consequently, serum triglyceride values were not determined, and instead of the Friedewald formula16 being used to calculate LDL cholesterol values, non-HDL cholesterol values are presented.
Children in whom baseline and at least one later lipid measurements were successfully obtained were included in the statistical analyses. Thus, the cholesterol analyses comprise 748 children (70% of the whole trial cohort), and the apolipoprotein analyses include 664 children.
The results are shown as mean±SD values with 95% CIs for the mean. Serum lipid values were normally distributed. To selectively analyze the effect of intervention, values were adjusted for the baseline situation (7-month lipid values). Longitudinal data of serum lipid concentrations were analyzed with unbalanced repeated-measures ANCOVA17 with the 7-month lipid value as covariate. The differences were considered significant at P<.05. A two-sample t test was used in comparison of dietary data of the two study groups. SAS release 6.08 program package (SAS Institute) and BMDP version 1990 (BMDP Statistical Software) were used.
The required sample size for the trial was predicted to achieve, at a 1% significance with 80% power, a 0.2-mmol/L true difference in the change of serum cholesterol concentration between the study groups, assuming that the SD of serum cholesterol concentration is 0.9 mmol/L.
Baseline nutrient intakes of the intervention and control groups (children with formula as their only milk source) at 8 months of age were similar (Tables 1 and 2⇓⇓). Daily energy intake was slightly lower in the intervention children than in control children during the trial. The intervention children, irrespective of sex, consumed less fat and saturated fat than the control children, and their relative intakes of fat (E%) and cholesterol were also lower throughout the intervention. The intervention children had higher intakes of polyunsaturated fats and consistently higher P/S ratios in their diet than the control children (P<.0001 to P=.031 throughout the follow-up).
Boys always had higher mean intakes of energy than girls in both groups of children (P<.0001 to P=.044 for the difference between sexes). Boys always consumed slightly more fat than girls (P=.0017 to P=.54). The intervention girls had higher relative intakes of saturated fat (E%) than the intervention boys at 13 and 24 months of age (P=.0035, 95% CI of the difference between sexes, 0.3 to 1.7 E% at 13 months; P=.0049, 95% CI, 0.0 to 1.3 E% at 24 months), and similarly, their relative cholesterol intake was higher at 13 months (P=.0081, 95% CI, 3 to 19 mg/1000 kcal). Relative intakes of saturated fat and cholesterol were always similar in the boys and girls of the control group.
Lipids and Lipoproteins
All serum lipid and lipoprotein concentrations and their ratios were similar in the intervention and control children at 7 months of age. Baseline serum cholesterol (P<.0001), non-HDL cholesterol (P<.0001), and apo B (P<.0001) differed between the sexes, with girls having higher values than boys. The ratios of HDL cholesterol to cholesterol (P=.0002) concentration and apo A-I to apo B (P<.0001) concentration differed also, but boys had higher values than girls.
From 13 to 36 months, the baseline adjusted mean serum cholesterol concentration was lower in the intervention children than in the control children (P<.0001 for the group effect; Figure⇓). The difference between the adjusted means of the two groups of children (intervention versus control) remained stable during this period (P=.28 for groups by time interaction). The 95% CI for the difference of the baseline adjusted means was from −0.27 to −0.12 mmol/L.
Similarly, the baseline adjusted means of serum HDL cholesterol, non-HDL cholesterol, apo A-I, and apo B concentrations were lower in the intervention group (P<.0001, P<.0001, P<.0001, and P=.021 for group effect, respectively), and the difference between the means did not change during follow-up (P=.60, P=.29, P=.56, and P=.99 for groups by time interaction, respectively). The 95% CI for the baseline adjusted mean difference between the intervention and control children was −0.07 to −0.03 mmol/L for HDL cholesterol, −0.22 to −0.08 mmol/L for non-HDL cholesterol, −0.05 to −0.02 g/L for apo A-I, and −0.04 to −0.003 g/L for apo B. Baseline adjusted means of the ratios of HDL cholesterol to total cholesterol and of apo A-I to apo B did not differ between the intervention and control children (P=.59 and P=.38 for the group effect, respectively).
Because of the interaction between sex and treatment group in serum cholesterol (P=.011) and non-HDL cholesterol (P=.017), the sexes were analyzed separately (Tables 3 and 4⇓⇓). Baseline adjusted mean serum lipid concentrations differed between intervention and control boys (P<.0001) but not between intervention and control girls (P=.089). The 95% CI for the difference of adjusted means of serum cholesterol was from −0.39 to −0.20 mmol/L in boys during follow-up, whereas it was from −0.21 to 0.01 mmol/L in girls. The baseline adjusted mean cholesterol, non-HDL cholesterol, and HDL cholesterol concentrations at 36 months were ≈6% higher in the control boys than in the intervention boys and 3% higher in the control girls than in the intervention girls. Girls had higher serum cholesterol concentrations than boys (P=.046) throughout the trial.
Growth of the children was not adversely influenced by the intervention (data not shown).
The randomized, prospective STRIP baby trial comprised more than 1000 healthy children, whose nutrient intakes, serum lipid values, growth, and development were regularly monitored. This trial clearly shows that an intervention aiming at decreased intake of saturated fat and cholesterol markedly influences serum lipid values and that children's lipid values can be retained at a low level without causing meaningful untoward effects in the children or their families. A benefit of ≈3% to 6% in serum cholesterol concentration persists at least up to the age of 3 years when families receive individualized counseling at regular intervals. Although serum lipid and lipoprotein values of the intervention children were, in the mean, significantly lower than those of the control children, they remained well within the age-adjusted reference values of healthy children. Furthermore, the growth of the children was measured frequently during the trial, and any deviations in growth led to even more detailed evaluation of dietary intakes and exclusion of other causes of growth retardation. Interestingly, the effect of intervention on serum cholesterol, non-HDL cholesterol, and HDL cholesterol was significant only in the male children. This sex difference may result in part from the slightly higher P/S ratio and lower relative saturated fat and cholesterol intake in the intervention boys than girls.
The intervention families were advised to change the child's milk type from breast milk or formula straight to skim (or 1% fat) milk at 1 year of age and to replace the “missing” milk fat with vegetable oil or soft margarine so as not to reduce intake of total fat or energy. However, against the plan, the intervention children consistently consumed slightly less fat and energy than the control children. Saturated fat and cholesterol intakes were lower in the intervention children, and their P/S ratio was higher. Dietary counseling was successful, since the diet of the intervention children changed and remained different from the diet of the control children. However, the P/M/S ratio of 1:1:1 in the diet was not reached.
Previously, Friedman and Goldberg18 showed that serum cholesterol concentration was 6% lower in 3-year-old children who had been on a low-saturated-fat, low-cholesterol diet since birth than in their control peers who had received an unrestricted diet. Serum cholesterol concentration decreased by 15% and HDL cholesterol concentration by 17% when 36 healthy 8- to 18-year-old children in eastern Finland consumed a diet in which the P/S ratio had been increased from 0.18 to 0.61 for 12 weeks.19 In hypercholesterolemic children, serum cholesterol concentrations have diminished at most by 10% to 15% by dietary intervention.20 21 In the present trial, healthy children not affected by hypercholesterolemia showed a long-term 3% to 6% reduction in serum cholesterol values compared with control children. Recently, in the DISC study,22 mean serum cholesterol concentration decreased significantly (7% to 8%) in 8- to 10-year-old hypercholesterolemic children in both intervention and usual-care groups. A large part of the reduction may be explained by the age of the subjects, because many of them were followed through puberty, which regularly induces marked decreases in serum cholesterol concentrations. In the DISC study, as in all intention-to-treat studies, there may have been some nonresponders in the intervention group. Meanwhile, some control children may have changed their dietary habits, because they also knew that their lipid values were high. In the DISC study, the net effect of the intervention on serum cholesterol concentration was small (0.08 mmol/L, 2%, P=.02) at 3 years of intervention. In a long-term dietary trial like ours, some unintentional intervention probably always occurs in the control group, since the control families unquestionably are also exposed to a rather extensive intervention. Blood samples were drawn from the control children as often as from the intervention children, and the parents received the results immediately. Furthermore, the control families also recorded the children's food intake as often as the intervention families. Despite apparent intervention, the dietary records of the control children showed that the P/S ratio of their diet and intake of fat in E% (0.33 and 33 E%, respectively, at 2 years of age) paralleled quite well the values found in forty-six 1- to 2-year-old children in Helsinki in 1988 (P/S ratio of 0.33 when two children on vegetable oil supplementation were excluded, 33 E% from fats).23 The mean cholesterol concentration was 4.4 mmol/L in weaned 1-year-old infants24 and 5.0 mmol/L in 3-year-old children25 in Finland in the early 1980s. Current cholesterol levels of healthy Finnish children <3 years old, apart from our trial, are unknown; thus, comparison of the serum lipid values of the control children with those of the age-adjusted general population is not directly possible.
In this trial, serum cholesterol, HDL cholesterol, non-HDL cholesterol, apo A-I, and apo B concentrations all diminished equally in the intervention group compared with control children. In other words, serum lipoprotein fractions all decreased, but the ratios of HDL cholesterol to cholesterol as well as of apo A-I to apo B remained unchanged. Since HDL is involved in the reverse transport of cholesterol from tissues to liver and HDL cholesterol concentration is inversely related to the incidence of CHD,26 27 one may speculate that our trial also led to some unfavorable alterations in serum lipids. Low-fat diets and diets rich in polyunsaturated fatty acids are known to reduce serum HDL cholesterol concentration.28 Because of such dietary changes, certain religious denominations with restricted dietary intakes and trends toward vegetarianism29 show low serum HDL cholesterol concentrations and nevertheless a low incidence of CHD. Mean serum HDL cholesterol concentrations are also higher in boys in countries in which the population's serum cholesterol concentration is high and the CHD incidence is likewise high.30 Differences in ratios of HDL cholesterol to total cholesterol between countries are smaller, but the ratio is slightly lower in, for example, some Asian countries in which the CHD incidence is also low.30 The concentrations of total and HDL cholesterol may increase when a “Western” diet containing large proportions of foods of animal origin is consumed. Thus, the higher the level of LDL cholesterol, the more HDL cholesterol is needed. The decrease in serum HDL cholesterol in the intervention children may be simply a result of lower fat intake and lower serum LDL concentrations. Because LDL cholesterol concentration is probably the best lipid predictor of population differences in rates of CHD mortality, decreasing total cholesterol (and LDL cholesterol) concentration is possibly the most important aim in reducing the risk of CHD, although the simultaneous decrease in serum HDL cholesterol concentration may slightly diminish the effect of the prevention.
The diets consumed by families over the years are surprisingly consistent; for example, intakes of nutrients at the age of 2 years correlate significantly with intakes at the age of 4 years.31 In our trial, a marked intervention effect was achieved even during the first 6 months,8 and clearly, the families had adapted well to the dietary changes, because the difference in serum lipid concentrations between the intervention and control children was significant throughout the trial. However, after the age of 13 months, serum cholesterol concentration began to increase in the intervention children as well, almost in parallel with the values in the control children, but the absolute difference between the intervention and control children's serum lipid values remained almost unchanged. A slight dilution of the intervention effect was evident in the food records also. The P/S ratio slightly decreased and the intake of saturated fat (as E%) increased during the follow-up. Thus, continuous counseling appears necessary and new methods of counseling are needed if the goal is to keep cholesterol concentrations at a persistently low level.
An extensive meta-analysis3 of the major intervention studies in adults clearly shows that the benefit of reduction of serum cholesterol concentration strongly relates to the age at which the reduction is achieved. A 0.6-mmol/L (≈10%) decrease in serum cholesterol concentration at the age of 40 years was associated with a 54% decrease in the incidence of CHD, whereas a similar change at age 50 years led to a 39% decrease, at age 60 years to a 27% decrease, and at age 70 and 80 years to a 20% decrease in CHD incidence. If permanent, a 3% to 6% reduction in serum cholesterol concentration, as achieved in this trial as early as in childhood, probably will have a marked effect on the incidence of CHD in adulthood. However, it is apparent that the ability of the family to maintain the child's low serum cholesterol concentration throughout life needs to be further confirmed before a true estimation of the prognostic value of the results becomes possible.
In conclusion, dietary intervention, if begun in infancy and continued regularly throughout early childhood, decreases exposure of the children to high saturated fat and cholesterol intake and to high serum lipid concentrations.
Selected Abbreviations and Acronyms
|CHD||=||coronary heart disease|
|E%||=||percent of energy|
|STRIP||=||Special Turku Coronary Risk Factor Intervention Project for Babies|
|WHO||=||World Health Organization|
This study was supported by grants from the Mannerheim League for Child Welfare; the Finnish Cardiac Research Foundation; the Foundation for Pediatric Research, Finland; the Academy of Finland; the Yrjo¨ Jahnsson Foundation; the Juho Vainio Foundation; the Turku University Foundation; Chymos Ltd; the Raisio Group; and Van den Bergh Foods Co.
- Received January 22, 1996.
- Revision received March 28, 1996.
- Accepted April 15, 1996.
- Copyright © 1996 by American Heart Association
Law MR, Wald NJ, Thompson SG. By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ.. 1994;308:367-373.
National Cholesterol Education Program. Report of the Expert Panel on Blood Cholesterol Levels in Children and Adolescents. Bethesda, Md: US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Heart, Lung, and Blood Institute, 1991. NIH publication 91-2732.
Conclusions, Guidelines and Recommendations from the IUNS/WHO Workshop. Nutrition in the pediatric age group and later cardiovascular disease. J Am Coll Nutr.. 1992;11:S1-S2.
Rastas M, Seppa¨nen R, Knuts L-R, Karvetti R-L, Varo P, eds. Nutrient Composition of Foods. Helsinki, Finland: Publications of the Social Insurance Institution; 1993.
Knuts L-R, Rastas M, Haapala P, Seppa¨nen R, Karvetti R-L, Aro A. Nutrica: a computer program for calculation of food and nutrient intake. Proc Third Meeting of Eurofoods, Warsaw, Poland, May 24-27, 1987:16. Abstract.
Richmond W. Preparation and properties of a cholesterol oxidase from Nocardia sp. and its application to the enzymatic assay of total cholesterol in serum. Clin Chem.. 1973;19:1350-1356.
Siedel J, Schlumberger H, Klose S, Ziegenhorn J, Wahleweld AW. Improved reagent for the enzymatic determination of serum cholesterol. J Clin Chem Clin Biochem.. 1981;19:838-839.
Kostner GM. Enzymatic determination of cholesterol in high density lipoprotein fractions prepared by polyanion precipitation. Clin Chem.. 1976;22:695. Letter.
Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem.. 1972;18:499-502.
Crowder MJ, Hand DJ. Analysis of Repeated Measures. 1st ed. London, England: Chapman & Hall; 1990.
Friedman G, Goldberg SJ. An evaluation of the safety of a low-saturated-fat, low-cholesterol diet beginning in infancy. Pediatrics.. 1976;58:655-657.
Kallio MJT, Salmenpera¨ L, Siimes MA, Perheentupa J, Miettinen T. Exclusive breast-feeding and weaning: effect on serum cholesterol and lipoprotein concentrations in infants during the first year of life. Pediatrics.. 1992;89:663-666.
Gordon DJ, Probstfield JL, Garrison RJ, Neaton JD, Castelli WP, Knoke JD. High-density lipoprotein cholesterol and cardiovascular disease: four prospective American studies. Circulation.. 1989;79:8-15.
Pocock SJ, Shaper AG, Phillips AN. Concentrations of high density lipoprotein cholesterol, triglycerides, and total cholesterol in ischaemic heart disease. BMJ.. 1989;298:998-1002.
Fraser GE. Determinants of ischaemic heart disease in Seventh-Day Adventists: a review. Am J Clin Nutr. 1988;48(suppl 3):833-836.
Knuiman JT, West CE, Katan MB, Hautvast JGAG. Total cholesterol and high-density lipoprotein cholesterol levels in populations differing in fat and carbohydrate intake. Arteriosclerosis.. 1987;7:612-619.