Risk of CHD

Epidemiological Evidence. In populations that consume very low fat diets, such as those of rural Asia and Africa, the prevalence of CHD is low (Campbell et al., 1998; Singh et al., 1995; Tao et al., 1989; Walker and Walker, 1978). This fact has led to the concept that low fat diets will protect against CHD. However, this conclusion must be drawn with caution when it is applied to societies in which dietary and exercise habits differ markedly from societies in rural Asia and Africa. In the latter societies, people are highly active and lean (Singh et al., 1995; Walker and Walker, 1978). Both of these factors independently reduce risk for CHD and could offset any potentially detrimental effects of very low fat diets. For this reason, the effects of low fat diets must be viewed in the context of current societal habits in the United States and Canada and of changing habits in developing countries. Furthermore, in more recent years it has become clear that the relationship between fat intake and CHD is related more to the quality of fat than to the quantity. The relationship is clearly shown by cross-population studies. For example, some Mediterranean populations consume diets that are high in total fat and unsaturated fatty acids but low in saturated fatty acids; in these populations, rates of CHD are relatively low (Keys et al., 1980, 1984). In contrast, in northern Europe, where intakes of saturated fatty acids are high, so are rates of CHD (Keys et al., 1980, 1984). Two epidemiological studies showed no relationship between carbohydrate intake and LDL cholesterol concentration (Fehily et al., 1988; Tillotson et al., 1997).

In several recent, long-term prospective studies of diet and chronic disease, rates of CHD did not substantially differ across populations that consumed approximately 25 to 45 percent of energy from fat (Ascherio et al., 1996; Hu et al., 1997). Men who developed CHD were shown to consume a slightly higher percentage of energy from fat (34.7 percent) compared with those who did not develop CHD (33.3 percent); however, this small difference in fat intake may not be significant since intake was based on a


24-hour recall, and the data were not adjusted for energy intake (McGee et al., 1984). Furthermore, Hawaiians, who have a higher incidence of CHD than Japanese living in Hawaii, consumed more energy from fat (35 percent) than the Japanese (31 percent) (Bassett et al., 1969). It has been reported that those who developed CHD consumed slightly less energy from carbohydrate compared to those who did not develop CHD (Kushi et al., 1985; McGee et al., 1984) (Table 11-7). Other studies showed no significant association between risk of CHD and total carbohydrate or sugar intake (Bolton-Smith and Woodward, 1994; Liu et al., 1982, 2000).

Interventional Evidence. Increasing fat intake, as a result of increased saturated fat intake, has been shown to increase LDL cholesterol concentrations (Table 11-2), and therefore risk of CHD. Intervention studies that have investigated the effect of carbohydrate intake on LDL cholesterol concentration have shown mixed results (Table 11-3). Two intervention studies agree with the findings of West and colleagues (1990) in that LDL cholesterol concentration increased when the percent of energy from carbohydrate was decreased from 55 to 31 percent (Borkman et al., 1991) and 59 to 41 percent (Marckmann et al., 2000). However, in other studies in which saturated fatty acids have remained constant, varying the percentage of total fat was found to not alter the LDL cholesterol concentration (Garg et al., 1994; Grundy et al., 1988).

Yu-Poth and colleagues (1999) conducted a meta-analysis on 37 intervention studies that evaluated the effects of the National Cholesterol Education Program's Step I and Step II dietary interventions on various cardiovascular disease risk factors. Reductions in plasma total cholesterol and LDL cholesterol concentrations were significantly correlated with reductions in percentages of total dietary fat, but these also included a decrease in saturated fatty acids. Similarly, individuals who consumed the Dietary Approaches to Stop Hypertension diet, which contains 27 percent of energy from fat and only 7 percent of energy from saturated fat, had reduced total and LDL cholesterol concentrations (Obarzanek et al., 2001b). Singh and colleagues (1992) reported that mortality from CHD and other causes was significantly lower when patients with acute myocardial infarction were fed a reduced fat diet.

The increase in LDL cholesterol concentration observed with increased fat intake is due to the strong positive association between total fat and saturated fat intake and the weak association between total fat and poly-unsaturated fat intake (Masironi, 1970; Stamler, 1979). This association is also observed in Appendix Tables K-4, K-5, K-7, and K-8. As shown in many studies, saturated fatty acids raise LDL cholesterol concentrations (see Chapter 8), whereas unsaturated fatty acids do not. In fact, n-6 poly-unsaturated fatty acids reduce serum LDL cholesterol concentrations some-


what compared with carbohydrate (Hegsted et al., 1993; Mensink and Katan, 1992). The adverse effects of saturated fats are discussed in Chapter 8.

It has been postulated that a high fat intake predisposes to a pro-thrombotic state, which contributes to venous thrombosis, coronary thrombosis, or thrombotic strokes (Barinagarrementeria et al., 1998; Kahn et al., 1997; Salomon et al., 1999). Consumption of diets high in fat (42 or 50 percent) have been shown to increase blood concentrations of the prothrombotic markers, blood coagulation factor VII (VIIc), and activated factor VII (VIIa) (Bladbjerg et al., 1994; Larsen et al., 1997). The concentration of factor VII is associated with increased risk of CHD (Kelleher, 1992). Furthermore, a significant and positive association was found between the level of dietary fat and factor VIIc concentration (Miller et al., 1989).

Relation of Intakes of Saturated Fatty Acids and Total Fat. When fat is consumed in typical foods it contains a mixture of saturated, polyunsaturated, and monounsaturated fatty acids. Even when the content of saturated fatty acids in consumed fats is relatively low, the intakes of these fatty acids can be high with high fat intakes. For example, if all of the dietary fats consumed were low in saturated fatty acids (e.g., 20 percent of fat energy), a total fat intake of 35 percent of total energy would yield a saturated fatty acid intake of 7 percent of total energy. Consumption of a variety of dietary fats would likely result in an even higher percentage of saturated fatty acids. Thus, in practical terms, it would be difficult to avoid high intakes of saturated fatty acids for most persons if total fat intakes exceeded 35 percent of total energy. This fact is revealed by attempts to create a variety of heart-healthy menus (National Cholesterol Education Program, 2001). Moreover, data from CSFII show that with increased fat intake, there tends to be a greater increase in saturated fatty acid intake relative to polyunsaturated fatty acid intake (Appendix Tables K-4, K-5, K-7, K-8; Masironi, 1970; Stamler, 1979). It should be pointed out, however, that when replacing saturated fatty acid intake with carbohydrate, there is no effect on the total cholesterol:HDL cholesterol ratio (Mensink and Katan, 1992).

Conclusions. A few case-control studies have shown an association between total fat intake and risk for CHD. However, a detailed evaluation of these studies shows that it is not possible to separate total fat intake from saturated fatty acid intake, which is known to raise LDL cholesterol concentrations. Unsaturated fatty acids, which do not raise LDL cholesterol concentrations compared with carbohydrate, have not been implicated in risk for CHD through adverse effects on lipids or other risk factors. Nonetheless, practical efforts to create "heart-healthy" menus reveal that intakes of total fat exceeding 35 percent of total energy result in unacceptably high intakes oo o o

TABLE 11-7 Epidemiological Studies on Total Carbohydrate and Sugar Intake and Risk of Coronary Heart Disease (CHD)


Study Design



McGee et al., 1984

Multi-country bivariate analysis

7,088 men Prospective cohort, 10-y follow-up

Mean carbohydrate intake (% of energy) Non-CHD 46.5°

CHD 45.0b

No significant association between sugar intake and CHD

Those who developed CHD consumed less energy as carbohydrates No association between sugar intake and risk of CHD

Kushi et al., 1985

1,001 men Prospective cohort, 20-y follow-up

Mean carbohydrate intake (% of energy) No CHD death 42.7° CHD death 41.2b

Mean sugar intake (% of energy) No CHD death 17.3° CHD death 16.9°

Those who died from CHD consumed significantly less total carbohydrate No association between sugar intake and risk of CHD death


11,626 men and women

Mean sugar intake (% of energy)

No association between risk of CHD and

Smith and

Cross-sectional survey





either intrinsic or extrinsic sugar










Added sugar







Liu et al.,

75,521 women

Relative risk of CHD

No significant association between risk


Prospective cohort,



of CHD and total carbohydrate,

10-y follow-up

of intake

hydrate Sucrose


sucrose, or fructose intake





















a'b Within each study, the mean sugar or carbohydrate intakes that are significantly different between treatment groups have a different superscript.


of saturated fatty acids. Moreover, there is the possibility that high fat intakes may enhance a prothrombotic state, although the evidence to support this mechanism for enhancing CHD risk is not strong enough alone to make solid recommendations.

Risk of Hyperinsulinemia, Glucose Intolerance, the Metabolic Syndrome, and Type 2 Diabetes

The metabolic syndrome (insulin-resistance syndrome) describes a clustering of metabolic abnormalities including insulin resistance (with or without glucose intolerance), an atherogenic lipid profile (high triacylglycerol concentration, low HDL cholesterol concentration, and high small, dense LDL), raised blood pressure, a prothrombotic state, and a proinflammatory state (Reaven, 2001). A prothrombotic state is characterized by elevations of plasminogen activator inhibitor and high fibrinogen concentrations, whereas a proinflammatory state is indicated by high c-reactive protein concentrations and other inflammatory markers. Abdominal obesity (waist circumference > 102 cm in men and 88 cm in women) is highly correlated with the presence of insulin resistance (NHLBI/NIDDK, 1998) and is considered to be one of the clinical components of the metabolic syndrome (National Cholesterol Education Program, 2001). An excess of intra-abdominal fat has been identified as being highly associated with the lipid risk factors of the metabolic syndrome (Després, 1993), although total abdominal fat appears to be even more highly predictive of the insulin resistance component of the syndrome (Abate et al., 1996; Peiris et al., 1988). Many persons with the metabolic syndrome eventually develop type 2 diabetes. Thus, both obesity and weight gain are undisputed as major risk factors for the development of type 2 diabetes (defined as fasting plasma glucose > 7 mmol/L) (American Diabetes Association, 2001).

The contribution of diet per se to the development of type 2 diabetes is less clear. In some laboratory animals (e.g., some species of rodents), a high percentage of fat in the diet will induce insulin resistance (Budohoski et al., 1993; Chisholm and O'Dea, 1987). An important question is whether humans are similarly susceptible to this phenomenon independent of the effects of total fat intake on body fat content. Human studies do not provide a clear answer to this question. Thus, if higher intakes of total fat lead to obesity, this in and of itself will reduce insulin sensitivity and predispose to the metabolic syndrome and type 2 diabetes. Recent studies have demonstrated that reduced fat intake and weight loss result in improved glucose tolerance and reduced risk of type 2 diabetes (Swinburn et al., 2001; Tuomilehto et al., 2001).


Epidemiological Evidence. In several population studies, investigators have attempted to determine the contribution of total fat intake to either insulin sensitivity or diabetes. These analyses are difficult to interpret because of the multiplicity of potential confounding variables. Nevertheless, several studies have reported an association between higher fat intakes and insulin resistance as indicated by high fasting insulin concentration, impaired glucose tolerance, or impaired insulin sensitivity (Lovejoy and DiGirolamo, 1992; Marshall et al., 1991; Mayer et al., 1993), as well as to the development of type 2 diabetes (West and Kalbfleisch, 1971). A number of studies, however, have not shown this association (Coulston et al., 1983; Liu et al., 1983; Salmerón et al., 2001). In the Insulin Resistance Atherosclerosis Study, total fat intake univariately correlated with less insulin sensitivity (Mayer-Davis et al., 1997); however, in multiple regression analyses, the presence of obesity appeared to be a confounding variable. Lovejoy and DiGirolamo (1992) likewise found intercorrelations among insulin resistance, total fat intake, and obesity. In contrast, Larsson and coworkers (1999) found no evidence of independent effects of diet on insulin secretory or sensitivity among 74 postmenopausal women. Although several studies suggest an association between total fat intake and the presence of insulin resistance (Lovejoy, 1999; Vessby, 2000), the degree to which the relationship is mediated by obesity remains uncertain. Decreased physical activity is also a significant predictor of higher postprandial insulin concentrations and may confound some studies (Feskens et al., 1994; Parker et al., 1993).

Interventional Evidence. A number of metabolic and intervention studies have examined the relationships among fat intake, fasting glucose and insulin concentrations, areas under curves for plasma glucose and insulin concentrations, insulin sensitivity, glucose effectiveness, and glucose disposal rates (Table 11-8). Several studies reported that diets containing 35 percent fat were accompanied by more impaired glucose tolerance than diets containing 25 percent fat or less (Fukagawa et al., 1990; Jeppesen et al., 1997; Straznicky et al., 1999; Swinburn et al., 1991). Coulston and coworkers (1983) found that a diet containing 41 percent fat led to significantly higher concentrations of insulin in response to meals compared with a diet containing 21 percent fat, but there were no alterations in fasting concentrations. In other studies, no effect on measures of glucose tolerance were reported when diets varied in fat content from 11 to 30 (Leclerc et al., 1993) or 20 to 50 percent fat (Abbott et al., 1989; Borkman et al., 1991; Howard et al., 1991; Thomsen et al., 1999). When the diet was high in fat (50 percent of energy), the area under the curve for plasma glucose and insulin concentration was lower than when the diet had a low fat content (25 percent of energy) (Yost et al., 1998). In this study, the decreased


TABLE 11-8 Interventional Studies on the Effect of Dietary Fat on the Metabolic Parameters for Glucose and Insulin in Healthy Individuals


Study Design

Percent Fasting Fasting of Fat Glucose Insulin

Coulston et al., 1983

11 men and women 10-d crossover


Chen et al., 1988

8 young men 3- to 5-d crossover

42 55



10 elderly men 3- to 5-d crossover


Abbott et al., 1989

9 men and women 5-wk crossover

42-21 NSC

Fukagawa et al. 1990

6 young men 21-to 28-d intervention

42-14 Decreased* Decreased*

6 elderly men and women 21-to 28-d intervention

38-15 Decreased* Decreased ib

Borkman et al., 1991

8 men and women 3-wk crossover

20-50 NSC

Howard et al., 1991

7 men and women 5- to 7-wk crossover

42-21 NSC

9 men and women 3- to 5-wk longitudinal

12 Caucasians and 12 Pima Indians 2-wk crossover

42-21 NSC

15-50 Increased^ NSC


Area Under the Curve for Glucose

Area Under the Curve for Insulin

Insulin Sensitivity

Glucose Effectiveness

Glucose Disposal/ Disappearance Rate

Decreased b

ND ND Decreasedb NSC ND

ND ND Increasedb NSC ND

ND ND Decreasedb NSC ND


ND ND ND ND Increasedb





Increased^ Increased^ NSC Decreased^ ND



TABLE 11-8 Continued


Study Design

Percent of Fat

Fasting Glucose

Fasting Insulin

Swinburn et al.,

24 Caucasians and





Pima Indians 2-wk crossover

Leclerc et al.,

7 men and women





7-d crossover

Jeppesen et al., 1997

10 women 3-wk crossover




Yost et al., 1998

25 men and women 15-d crossover




Straznicky et al., 1999

14 men 2-wk crossover




Thomsen et al.,

16 men and women





4-wk crossover

Kasim-Karakas et al., 2000

54 postmenopausal women 4- to 12-mo crossover

15, 25, and 34



c ND = no data available. d p < 0.001. e p < 0.01.

fat intake was accompanied by an increased percentage of energy from carbohydrate. Garg and coworkers (1992b) reported that insulin sensitivity, indicated by insulin-mediated glucose disposal, was similar after almost a month of ingestion of either a reduced fat (25 percent of energy) or an increased fat diet (50 percent of energy). However, favorable effects of substituting a monounsaturated fat diet for a saturated fat diet on insulin sensitivity were seen at a total fat intake of up to 37 percent of energy (Vessby et al., 2001). A large, long-term intervention trial in adults showed that reducing total fat intake, in part, reduced the risk of the onset of type 2 diabetes by 58 percent (Tuomilehto et al., 2001). Similarly, the Diabetes Prevention Program Research Group reported that diet modification,


Area Under the Curve for Glucose

Area Under the Curve for Insulin

Insulin Glucose

Sensitivity Effectiveness

Glucose Disposal/ Disappearance Rate

Increased^ Increased^

Decreased d

Increasedd ND

Decreased^ Decreased^ ND

Increased^ NSC

Decreased^ ND

including a reduction of total fat intake from 34 to 27 percent of energy reduced the incidence of type 2 diabetes by 58 percent. Thus, there is no definitive evidence from metabolic and interventional studies that higher fat intakes impair insulin sensitivity in humans as they do in various laboratory animals. Any suggestive links between fat intake and either insulin secretion or sensitivity may be mediated through confounding factors, such as body-fat content, making it difficult to detect any independent contribution of total fat intake to insulin sensitivity.

Conclusions. Although high fat diets can induce insulin resistance in rodents, investigations in humans fail to confirm this effect. Moreover, an


association between dietary fat intake and risk for diabetes has been reported in some epidemiological studies, but this association is most likely confounded by various factors, such as obesity and glycemic index.

Risk of Cancer

High intakes of dietary fat have been implicated in the development of cancer, especially cancer of the lung, breast, colon, and prostate gland. Early support for this theory comes from laboratory animal and cross-cultural studies. The latter were based largely on international food disappearance data and migrant and time trend studies. In recent years, the theory that a diet high in fat predisposes to certain cancers has been weakened by additional epidemiological studies. Early cross-cultural and case-control studies reported strong associations between total fat intake and breast cancer (Howe et al., 1991; Miller et al., 1978; van't Veer et al., 1990), yet a number of epidemiological studies, most in the last 15 years, have found little or no association between fat intake and breast cancer (Hunter et al., 1996; Jones et al., 1987; Kushi et al., 1992; van den Brandt et al., 1993; Velie et al., 2000; Willett et al., 1987, 1992). A meta-analysis of 23 studies yielded a relative risk of 1.01 and 1.21 from cohort and case-control studies, respectively (Boyd et al., 1993).

Total fat intake in relation to colon cancer has strong support from animal studies (Reddy, 1992). However, evidence from epidemiological studies has been mixed (De Stefani et al., 1997b; Giovannucci et al., 1994; Willett et al., 1990). Howe and colleagues (1997) reported no association between fat intake and risk of colorectal cancer from the combined analysis of 13 case-control studies.

Epidemiological studies tend to suggest that dietary fat intake is not associated with prostate cancer (Ramon et al., 2000; Veier0d et al., 1997b). Giovannucci and coworkers (1993), however, reported a positive association between total fat consumption, primarily animal fat, and risk of advanced prostate cancer. Findings on the association between fat intake and lung cancer have been mixed (De Stefani et al., 1997a; Goodman et al., 1988; Veier0d et al., 1997a; Wu et al., 1994).

Risk of Nutrient Inadequacy or Excess

Diets High in Fat. With increasing intakes of carbohydrate, and therefore decreasing fat intakes, there is a trend towards reduced consumption of dietary fiber, folate, and vitamin C (Appendix K). With higher fat intakes, it is difficult to create practical high fat menus that do not contain unacceptably high amounts of saturated fatty acids (National Cholesterol Education Program, 2001).


Diets Low in Total Sugars. Micronutrient inadequacy can occur when sugars intake is very low (less than 4 percent of total energy) (Bolton-Smith and Woodward, 1995) because many foods that are abundant in micronutrients, such as fruits and dairy products, also contain naturally occurring sugars. A wide variety of foods from different food groups are needed to meet nutrient requirements. Because sugars are important for the palatability of foods, the complete omission of sugars from the diet could endanger overall nutrient adequacy by leading to low total energy intake, as well as low micronutrient intakes (Bolton-Smith, 1996). Although reduced nutrient intakes have been reported, adverse affects on health have not. Individuals with fructose intolerance, a condition caused by fructose-1-phosphate aldolase deficiency, strictly avoid foods containing fructose and sucrose and yet remain in good health (Burmeister et al., 1991).

AMDRs for Adults

When fat intakes are low and carbohydrate intakes are high, intervention studies, with the support of epidemiological studies, demonstrate a reduction in plasma HDL cholesterol concentration, an increase in the plasma total cholesterol:HDL cholesterol ratio, and an increase in plasma triacylglycerol concentration, which are all consistent with an increased risk of CHD. Conversely, many interventional studies show that when fat intake is high, many individuals consume additional energy, and therefore gain additional weight. Weight gain on high fat diets can be detrimental to individuals already susceptible to obesity and can worsen the metabolic consequences of obesity, particularly the risk of CHD. Moreover, high fat diets are usually accompanied by increased intakes of saturated fatty acids, which can raise plasma LDL cholesterol concentrations and further increase risk for CHD. Based on the apparent risk for CHD that may occur on low fat diets, and the risk for increased energy intake and therefore obesity with the consumption of high fat diets, the AMDR for fat and carbohydrate is estimated to be 20 to 35 and 45 to 65 percent of energy, respectively, for all adults. By consuming fat and carbohydrate within these ranges, the risk for obesity, as well as for CHD and diabetes, can be kept at a minimum. Furthermore, these ranges allow for sufficient intakes of essential nutrients while keeping the intake of saturated fatty acids at moderate levels.

There is no lower limit of intake and no known adverse effects with the chronic consumption of Dietary Fiber or Functional Fiber (Chapter 7). Therefore, an AMDR is not set for Dietary, Functional, or Total Fiber.


Maximal Intake Level for Added Sugars

Data from various national surveys show that increasing intakes of added sugars is associated with a decline in the consumption of certain micronutrients, thus increasing the prevalence of those consuming below the EAR or the AI. While such trends exist, it is not possible to determine a defined intake level at which inadequate micronutrient intakes occur. Furthermore, at very low or very high intakes, unusual eating habits most likely exist that allow for other factors to contribute to low micronutrient intakes. Based on the available data, no more than 25 energy from added sugars should be comsumed by adults. A daily intake of added sugars that individuals should aim for to achieve a healthy diet was not set. Total sugars intake can be lowered by consuming primarily sugars that are naturally occurring and present in micronutrient-rich foods, such as milk, dairy products, and fruits, while at the same time limiting consumption of added sugars from foods and beverages that contain minimal amounts of micro-nutrients, such as soft drinks, fruitades, and candies.

Low Fat, High Carbohydrate Diets of Children Fat Oxidation

Jones and colleagues (1998) reported a significantly greater fat oxidation in children (aged 5 to 10 years, n = 12) than in adults (aged 20 to 30 years, n = 6). Breath 13COg was measured in 12 children and 6 men following an oral bolus dose of [1-13C]palmitic acid (10 mg/kg of body weight) consumed with a test meal. Breath 13CO2 excretion was less in the men (35.1 percent of absorbed dose, P = 0.005) than in the children (57.0 percent of absorbed dose). The children exhibited greater fat oxidation in the postabsorptive state (2.43 g/h) and postprandial (11.89 g/6 h) states than the men (0.93 g/h postabsorptive, 9.86 g/6 h postprandial). The children also had greater fat oxidation compared with women studied previously by these investigators (0.53 g/h postabsorptive, 0.03 g/6 h postprandial) (Murphy et al., 1995).


Most studies have reported no effect of the level of dietary fat on growth when energy intake is adequate (Boulton and Magarey, 1995; Fomon et al., 1976; Lagstrom et al., 1999; Lapinleimu et al., 1995; Niinikoski et al., 1997a, 1997b; Obarzanek et al., 1997; Shea et al., 1993). Two well-controlled trials demonstrated that a diet providing less than 30 percent energy from fat does not result in adverse effects on growth in


children up to 8 years of age (Lapinleimu et al., 1995; Niinikoski et al., 1997a, 1997b). A cohort study with a 25-month follow-up showed that there was no difference in stature or growth of children aged 3 to 4 years at baseline across quintiles (27 to 38 percent) of total fat intake (Shea et al., 1993). The Special Turku Coronary Risk Factor Intervention Project showed no difference in growth of children 7 months to 5 years of age when they consumed 21 to 38 percent fat (Lagstrom et al., 1999). Niinikoski and coworkers (1997a) reported that 1-year-old children who consistently consumed low fat diets (less than 28 percent) grew as well as children with higher fat intakes. A cohort study showed that children aged 2 years in the lower tertile of fat intake (less than 30 percent) had a height and weight similar to that of the higher fat intake groups (Boulton and Magarey, 1995).

A few studies have observed impaired growth among hypercholsterolemic children who were advised to consume 30 percent or less of energy from fat. However, the energy intake was also reduced (Lifshitz and Moses, 1989) or not reported (Hansen et al., 1992). In a group of Canadian children 3 to 6 years of age, a fat intake of less than 30 percent of energy was associated with an odds ratio of 2.3 for weight-for-age below the 50th percentile at 6 years of age (Vobecky et al., 1995). A comprehensive evaluation of the effect of diet-related variables on the growth of children under 6 years of age from 18 Latin American countries (FAO/WHO, 1996) demonstrated that diets providing less than 22 percent energy from fat and with less than 45 percent of total fat from animal fat were related to low birth weight, underweight, and stunting (height-for-age < 2 standard deviations) (Uauy et al., 2000). The dietary determinants that best explained low birth weight were energy, protein, and animal fat, suggesting that high-quality animal protein and associated nutrients are important for growth and development.

Risk of Nutrient Inadequacy or Excess

Diets High in Carbohydrate and Low in Fats. Because the diets of young children are less diversified than that of adults, the risk of inadequate micronutrient intake is increased in these children. A cohort of 500 children aged 3 to 6 years showed that those who consumed less than 30 percent of energy from fat consumed less vitamin A, vitamin D, and vitamin E compared with those who consumed higher intakes of fat (30 to 40 percent) (Vobecky et al., 1995). Calcium intakes decreased by more than 100 mg/d for 4- and 6-year-old children who consumed less than 30 percent of energy from fat (Boulton and Magarey, 1995). Lagstrom and coworkers (1997, 1999), however, did not observe reduced intakes of micronutrients in children with low fat intakes (26 percent).


The Dietary Intervention Study in Children (DISC), a multi-center, randomized trial of children 8 to 10 years of age, demonstrated that reducing the intake of fat to 28 percent of energy over a 3-year period increased the percentage of children not meeting the RDA for vitamin E and zinc; however, no biochemical evidence of deficiency of these nutrients was found (Obarzanek et al., 1997). Tonstad and Sivertsen (1997) observed no reduced intake of micronutrients with diets providing 25 percent of energy as fat. Nicklas and coworkers (1992) reported reduced intakes of certain micronutrients by 10-year-old children who consumed less than 30 percent of energy as fat; however, this level of fat intake was associated with marked increased intakes of candy. It has been suggested that children who consume a low fat diet can meet their micronutrient recommendation by appropriate selection of certain low fat foods (Peterson and Sigman-Grant, 1997). This is especially true for older children whose diets are typically more diverse.

The tables in Appendix K show the intakes of nutrients at various intake levels of carbohydrate. With increasing intakes of carbohydrate, and therefore decreasing intakes of fat, the intake levels of calcium and zinc markedly decreased in children 1 to 18 years of age (Appendix Tables K-1 through K-3).

Diets High in Added Sugars. Several surveys have evaluated the impact of added sugars intake on micronutrient intakes in children (Table 11-5). Gibson (1997) examined data from the U.K. National Diet and Nutrition Survey of Children Aged 1.5 to 4.5 Years (boys, n = 848; girls, n = 827) and found evidence of a nutrient dilution effect by nonmilk extrinsic sugars (NMES). Children consuming the highest concentrations of NMES (greater than 24 percent of energy) had intakes of most micronutrients that were between 6 and 20 percent below average. Gibson (1997) concluded that the inverse association of NMES with micronutrient intakes was of most significance for the 20 percent of children with the diets highest in NMES (24.9 percent of energy for boys and 24.5 percent of energy for girls).

In a study of British adolescents, reduced intakes of calcium, phosphorus, iron, vitamin A, vitamin D, and folic acid were associated with increased sugars intakes (mean added sugars intake for the high sugars consumers was 122 g/d for boys and 119 g/d for girls) (Rugg-Gunn et al., 1991). In a smaller survey (n = 143), added sugars intakes at levels as high as 27 percent of energy did not have a significant impact on micronutrient intakes (Nelson, 1991).

Similar to that observed for adults using data from NHANES III, increasing the added sugars intake by every 5th percentile tended to be associated with reduced intakes of certain micronutrients, including


calcium, vitamin A, iron, and zinc (Appendix Tables J-1 through J-3, J-6, and J-7). This reduction in micronutrient intake was most significant when added sugars intake levels exceeded 25 percent of energy.

From 1989 to 1995, energy intakes increased for U.S. children aged 2 to 17 years primarily due to increased carbohydrate consumption. Beverages, particularly soft drinks, were important contributors to the increased carbohydrate consumption. During this period, micronutrient intakes (except for iron) did not increase and calcium intakes decreased. This was attributed to the fact that increased energy was largely obtained from soft drinks, which do not add nutrients and displace milk in children's diets, with negative consequences for total diet quality (Morton and Guthrie, 1998).

Children who were high consumers of nondiet soft drinks had lower intakes of riboflavin, folate, vitamin A, vitamin C, calcium, and phosphorus in comparison with children who were nonconsumers of soft drinks (Harnack et al., 1999). Several of these nutrients (folate, vitamin A, and calcium) have been identified in national surveys as "shortfall" or "problem" nutrients among various age and gender groups (ARS, 1998). Ballew and colleagues (2000) demonstrated that in U.S. children, milk consumption was positively associated with the likelihood of achieving recommended vitamin A, vitamin B12, folate, calcium, and magnesium intakes in all age groups. Juice (100 percent fruit or vegetable juice) consumption was positively associated with achieving vitamin C and folate recommended intakes in all age groups, as well as magnesium intake among children aged 6 years and older. Soft drink intake was negatively associated with achieving recommended vitamin A intake in all age groups, calcium in children younger than 12 years of age, and magnesium in children 6 years of age and older.

Others have shown that children who consumed milk at the noon meal had the highest daily intakes of vitamin A, vitamin E, calcium, and zinc, whereas the opposite was true for children who consumed soft drinks and tea (Johnson et al., 1998). Hence, beverages that are major contributors of the naturally occurring sugars, such as lactose and fructose, in the diet (e.g., milk and fruit juice) have been positively associated with nutrient adequacy, while beverages that are the principal source of added sugars in the diet (e.g., soft drinks) have been negatively associated with nutrient adequacy in the diets of U.S. children and adolescents (Ballew et al., 2000; Johnson et al., 1998).

Diets High in Total Sugars. The findings from three surveys on the relationship between total sugars intake and micronutrient intake in children are mixed (Table 11-6). Gibson (1993) did not observe reduced micronutrient intakes when total sugars intake exceeded 25 percent of energy. Nicklas and coworkers (1996) reported that the percent of children meeting the RDA for only niacin and zinc was significantly reduced


when the intake of total sugars exceeded 31 percent of energy. A linear reduction in several micronutrients was observed with increasing total sugars intake (Farris et al., 1998).

High Fat, Low Carbohydrate Diets of Children Risk of Obesity

In the United States and Canada, there is evidence that children are becoming progressively overweight (Flegal, 1999; Gortmaker et al., 1987; Tremblay and Willms, 2000; Troiano et al., 1995). Furthermore, Serdula and coworkers (1993) reviewed a number of longitudinal studies with varying cut-off levels for obesity and concluded that 26 to 41 percent of obese preschool children and 42 to 63 percent of obese school-age children became obese adults. Clinical evidence of disease associated with excess body weight, reduced physical activity, or high dietary fat intakes, however, are generally absent. The evidence for a role of dietary fat intakes in promoting higher energy intakes and thus promoting obesity in young children is conflicting.

A positive trend in energy intake was associated with an increased percent of energy from fat for children up to 8 years of age (Boulton and Magarey, 1995). A positive correlation between fat intake and fat mass has been reported for boys 4 to 7 years of age (Nguyen et al., 1996). A lack of effect of dietary fat on BMI and adiposity, however, has been reported for children 1.5 to 4.5 years of age (Atkin and Davies, 2000; Davies, 1997).

The DISC trial found no difference in BMI for children 8 to 10 years of age who consumed diets containing 29 or 33 percent fat over a 3-year period (Lauer et al., 2000). However, several studies showed a positive correlation between dietary fat intake and body fatness in children 8 to 12 years of age (Maffeis et al., 1996; Obarzanek et al., 1994; Ricketts, 1997). The average fat intake of nonobese children was measured to be 31 to 34 percent for children 9 to 11 years old, whereas the average fat intake of obese children was 39 percent of energy (Gazzaniga and Burns, 1993). A positive association between fat intake and several adiposity indices were observed, but only for up to 35 percent of energy (Maillard et al., 2000). Other factors that have been associated with increased BMI include physical activity.

Risk of CHD

Clinical studies have provided some evidence that serum cholesterol concentration is modified in children the same way as in adults, with serum total, LDL, and non-HDL cholesterol concentrations being increased by


consuming diets higher in total fat (Lauer et al., 2000; Niinikoski et al., 1996; Obarzanek et al., 2001a; Shannon et al., 1994; Simell et al., 2000; Vartiainen et al., 1986). However, no significant association between dietary fat and LDL cholesterol concentration was observed for boys and girls (aged 8 to 10 years) consuming fat ranging from 10 to 50 percent of energy (R = -0.04 to 0.14) (Kwiterovich et al., 1997). Furthermore, a significant positive association between fat intake and total cholesterol concentration was observed in only two of five countries (Knuiman et al., 1983).

Another potential indicator for children's future risk of CHD is the presence of fatty streaks, which are found in the aortas of almost all children over 3 years of age in North America (Holman et al., 1958), and begin to appear in the coronary arteries about 5 to 10 years later than in the aorta (Berenson et al., 1992; McGill, 1968; Stary, 1989; Strong et al., 1992). The prevalence of aortic fatty streaks differs only slightly among children and adolescents of all populations studied, regardless of the frequency of atherosclerosis and coronary artery disease in adults of the respective population (Holman et al., 1958; McGill, 1968). The absence of a relation between aortic fatty streaks and the clinically relevant lesions of atherosclerosis in epidemiological and histological studies has thus raised questions on the clinical significance of fatty streaks in the aorta of young children (Newman et al., 1995; Olson, 2000). The Pathobiological Determinants of Atherosclerosis in Youth Study, however, has provided evidence that an unfavorable lipoprotein pattern (i.e., elevated non-HDL cholesterol and low HDL cholesterol concentrations), obesity, and hyperglycemia are associated with raised fatty streaks in the coronary artery and abdominal aorta in late teenage years (McGill et al., 2000a, 2000b). Similarly, the Bogalusa Heart Study observed a positive association between LDL cholesterol concentration and the percentage of surface with fatty streaks in the aorta (Berenson et al., 1992). These findings are consistent with the hypothesis of the progression of fatty streaks to fibrous plaques under the influence of the prevailing risk factors for coronary artery disease (McGill et al., 2000a, 2000b).

It is still unclear, however, how reduction in serum cholesterol concentration in childhood, if maintained, is associated with risk of CHD in adulthood. In addition, there are still pivotal issues that must be examined further, including the relationship between fatty streaks found in the arteries of young children and the later appearance of raised lesions associated with coronary vascular disease, the effects of dietary total fat modification on predictive risk factors in children, the safety of the diet with respect to total energy and micronutrients for the general population, and the long-term health benefit of establishing healthy dietary patterns early in childhood.


Risk of Nutrient Inadequacy or Excess

Appendix Tables K-1 through K-3 and K-6 provide data from CFSII on the intake of various nutrients based on the level of carbohydrate intake. It can been seen from these tables that as the level of carbohydrate intake decreases, and therefore the level of fat increases, certain nutrients such as folate and vitamin C markedly decrease. Furthermore, with increasing levels of fat intake, the intake of saturated fat relative to linoleic acid intake markedly increases.

AMDRs for Children

The evidence suggests that children have a higher fat oxidation rate compared to adults, and that reduced intake of certain micronutrients can occur with the consumption of low fat diets, whereas there is potential risk of obesity with high fat intakes. High intakes of fat may promote increased risk for CHD and obesity. Dietary fat provides energy, which may be important for younger children with reduced food intakes, particularly during the transition from a diet high in milk to a mixed diet. Thus, there should be a transition from the high fat intake during infancy (55 and 40 percent of energy for the 0- to 6- and 7- to 12-months age groups, respectively) (Chapter 8) to an AMDR for adults (20 to 35 percent of energy). Therefore, it is estimated that the AMDR for fat intake is approximately 30 to 40 percent of energy for children 1 to 3 years of age and 25 to 35 percent of energy for children 4 to 18 years of age. The AMDR for carbohydrate is the same as for adults (45 to 65 percent of energy). The ranges of fat intake include intakes of saturated fat that should be consumed at levels as low as possible while consuming a nutritionally adequate diet.

Maximal Intake Level for Added Sugars

As for adults, no more than 25 percent of energy from added sugars should be consumed by children to ensure adequate micronutrient intakes. For those children whose intake is above this level, added sugars intake can be reduced by consuming sugars that are primarily naturally occurring and present in foods such as milk, dairy products, and fruits, which also contain essential micronutrients.


Approximately 20 to 40 percent of fat is consumed as n-9 mono-unsaturated fatty acids, almost all of which is oleic acid (Appendix Tables E-1 and E-8). Monounsaturated fatty acids are not essential fatty acids, but they may have some benefit in the prevention of chronic disease. Although


early research pointed to this potential benefit, most attention has been given to it in the past decade.

Low n-9 Monounsaturated Fatty Acid Diets Risk of CHD

Epidemiological Evidence. Population data on monounsaturated fatty acid intake and risk of coronary heart disease (CHD) are limited. However, in long-term follow-up studies of the Seven Countries Study, higher intakes of monounsaturated fatty acids were associated with decreased rates of CHD mortality (Keys et al., 1986). Other reports indicate that monounsaturated fatty acids have a neutral or beneficial effect on risk (Hu et al., 1997; Kromhout and de Lezenne Coulander, 1984; Pietinen et al., 1997).

Interventional Evidence. Much work has been conducted and is ongoing to identify the ideal substitute for saturated fat in a blood cholesterol-lowering diet. The effects of a high monounsaturated fatty acid versus a low fat, high carbohydrate diet on serum lipid and lipoprotein concentrations have been a focus of considerable scientific inquiry. Eighteen well-controlled clinical studies that compared the effects of substituting mono-unsaturated fatty acids versus carbohydrate for saturated fat in a blood cholesterol-lowering diet have recently been reviewed (Kris-Etherton et al., 2000). In these studies, when on both high monounsaturated fat and low fat, high carbohydrate diets, saturated fatty acids contributed to 4 to 12 percent of energy and dietary cholesterol varied from less than 100 up to 410 mg/d. Diets high in monounsaturated fatty acids provided 17 to 33 percent of energy from monounsaturated fatty acids and contained more total fat (33 to 50 percent energy) than the low fat, high carbohydrate diets (18 to 30 percent energy). The low fat, high carbohydrate diets provided 55 to 67 percent of energy from carbohydrate. Compared to baseline values, serum total cholesterol concentrations changed from -17 to +3 percent on the low fat, high carbohydrate diet, whereas it changed from -20 to -3 percent on the high monounsaturated fatty acid diet. The range of decrease in plasma low density lipoprotein (LDL) cholesterol concentration was similar (-22 to +1 percent) among individuals on the two diets. The change in serum triacylglycerol concentrations ranged from -23 to +37 percent for individuals consuming the low fat, high carbohydrate diets and from -43 to +12 percent for diets high in monounsaturated fatty acids. Changes in high density lipoprotein (HDL) cholesterol concentrations ranged from -25 to +2 percent for individuals on the low fat, high carbohydrate diets compared to a -9 to +6 percent change for individuals on diets high in monounsaturated fatty acids.


These data indicate that in weight-stable individuals, a high mono-unsaturated fatty acid, low saturated fatty acid diet results in a more favorable metabolic profile with respect to total cholesterol, HDL cholesterol, and triacylglycerol concentrations. Figure 11-4 shows that with increased monounsaturated fatty acid intake, there is a favorable reduction in the total cholesterol:HDL cholesterol ratio. Furthermore, a meta-analysis of feeding studies estimated that the regression coefficients for the effects of monounsaturated fatty acids on LDL and HDL cholesterol concentrations were -0.008 and +0.006, respectively, suggesting a slight positive benefit (Clarke et al., 1997).

251 20-

Dietary MUFA (% energy)

FIGURE 11-4 Relationship between monounsaturated fatty acid (MUFA) intake and total cholesterol (TC):high density lipoprotein cholesterol (HDL-C) concentration ratio. Weighted least-squares regression analyses were performed using the mixed procedure to test for differences in lipid concentrations (SAS Statistical package, version 8.00, SAS Institute, Inc., 1999).

DATA SOURCES: Berry et al. (1992); Curb et al. (2000); Garg et al. (1988, 1992a, 1994); Ginsberg et al. (1990); Grundy (1986); Grundy et al. (1988); Jansen et al. (1998); Kris-Etherton et al. (1999); Lefevre et al., unpublished; Lopez-Segura et al. (1996); Mensink and Katan (1987); Nelson et al. (1995); Parillo et al. (1992); Pelkman et al. (2001); Perez-Jimenez et al. (1995, 1999, 2001).


Risk of Diabetes

Epidemiological studies tend to suggest no association between mono-unsaturated fatty acid intake and risk of indicators for diabetes (Feskens et al., 1995; Marshall et al., 1997). Similarly, some intervention studies showed no effect of monounsaturated fatty acid intake on indicators for risk of diabetes (Fasching et al., 1996; Roche et al., 1998; Thomsen et al., 1999; Vessby et al., 2001). Uusitupa and coworkers (1994), however, reported a significantly lower area under the curve for plasma glucose concentration and a greater glucose disappearance rate when healthy women consumed a diet rich in monounsaturated fatty acids (19 to 20 percent) compared with a diet rich in saturated fatty acids.

Risk of Cancer

Bartsch and colleagues (1999) reported a protective effect of oleic acid on cancer of the breast, colon, and possibly the prostate. A few epide-miological studies have reported an inverse relationship between mono-unsaturated fatty acid intake and risk of breast cancer (Willett et al., 1992; Wolk et al., 1998), while a number of studies reported no association (Holmes et al., 1999; Hunter et al., 1996; Jones et al., 1987; Kushi et al., 1992; van den Brandt et al., 1993; van't Veer et al., 1990). Increased consumption of olive oil was associated with significantly reduced breast cancer risk (La Vecchia et al., 1995; Martin-Moreno et al., 1994; Trichopoulou et al., 1995).

A diet high in monounsaturated fatty acid-rich vegetable oils, including olive, canola, or peanut oils, has been associated with a protective effect or no risk of prostate cancer (Norrish et al., 2000; Ramon et al., 2000; Schuurman et al., 1999; Veier0d et al., 1997b). Some speculate that the apparent protective effects of olive oil (and other vegetable oils) reflect constituents other than monounsaturated fatty acids including squalene (Newmark, 1999), phenolic compounds, antioxidants, and other compounds (Owen et al., 2000).

No significant association has been reported for monounsaturated fatty acid intake and risk of colorectal cancer (Giovannucci et al., 1994; Howe et al., 1997).

Risk of Nutrient Inadequacy

In the United States, monounsaturated fatty acids provide 12 to 13 percent of energy intake. About 50 percent of these fatty acids are consumed via animal products, primarily meat fat (Jonnalagadda et al., 1995). Vegetable oils that are good sources of monounsaturated fatty acids include canola


oil and olive oil. Although the major sources of monounsaturated fatty acids (animal fat and vegetable oils) are not required to supply essential nutrients, very low intakes of monounsaturated fatty acids would require increased intakes of other types of fatty acids to achieve recommended fat intakes. Consequently, intakes of saturated and n-6 polyunsaturated fatty acids would probably exceed a desirable level of intake (see "n-6 Polyunsaturated Fatty Acids" and Chapter 8).

High n-9 Monounsaturated Fatty Acid Diets

There are limited data on the adverse health effects from consuming high levels of n-9 monounsaturated fatty acids (see Chapter 8, "Tolerable Upper Intake Levels").

Acceptable Macronutrient Distribution Range n-9 Monounsaturated fatty acids are not essential in the diet, and the evidence relating low and high intakes of monounsaturated fatty acids and chronic disease is limited. Therefore, an Acceptable Macronutrient Distribution Range (AMDR) for n-9 monounsaturated fatty acids is not provided. Nonetheless, practical limits on intakes of monounsaturated fatty acids will be imposed by AMDRs for total fat and other types of fatty acids.


Low n-6 Polyunsaturated Fatty Acid Diets

Risk of CHD

Epidemiological Evidence. Many populations of the world, such as in Crete and Japan, have low total intakes of n-6 polyunsaturated fatty acids (e.g., < 4 percent of total energy) without obvious health consequences (Okita et al., 1995; Renaud et al., 1995). However, high intakes of n-6 polyunsaturated fats have been associated with blood lipid profiles (e.g., reduced total and low density lipoprotein [LDL] cholesterol, reduced triacylglycerol, and increased high density lipoprotein [HDL] cholesterol concentrations) that are associated with low risk of coronary heart disease (CHD) (Arntzenius et al., 1985; Becker et al., 1983; Sonnenberg et al., 1996). Prospective epidemiological evidence suggests that after controlling for other components of the diet, replacing saturated fats with unsaturated fats decreases risk of CHD (Hu et al., 1997); however, the dose-response


relationship between n-6 fatty acids and risk of CHD was not adequately established with certainty. An inverse association between linoleic acid intake and risk of coronary death was observed in several prospective studies (Arntzenius et al., 1985; Gartside and Glueck, 1993), while Pietinen and coworkers (1997) did not observe a relationship between linoleic acid intake and risk of CHD. A cross-sectional study showed that linoleic acid was inversely related to the prevalence of CHD, and this effect was stronger with higher intakes of linolenic acid (Djoussé et al., 2001). It is difficult to provide a direct assessment of n-6 fatty acids on risk of CHD without taking into consideration the impact of several dietary and nondietary factors, in addition to serum cholesterol concentrations, that lead to CHD and may be modified by the intake of saturated fat and n-6 fatty acids.

Interventional Evidence. From the standpoint of blood lipid concentration and CHD, higher n-6 polyunsaturated fatty acid intake generally alters blood lipid concentration to result in a decreased risk profile (Katan et al., 1994) (Table 11-9). Controlled trials have examined the effects of substituting n-6 fatty acids in the diet to replace carbohydrate or saturated fatty acids (Mensink et al., 1992). In general, any fat that replaces carbohydrate in the diet raises HDL cholesterol and decreases triacylglycerol concentrations, with only small differences between individual fatty acids. n-6 Fatty acids decrease LDL cholesterol concentrations to a much greater degree than do saturated fatty acids (Mensink et al., 1992).

Risk of Diabetes

A number of epidemiological studies have been conducted to ascertain whether the quality of fat can affect the risk for diabetes. An inverse relationship was reported for vegetable fats and polyunsaturated fats and risk of diabetes (Colditz et al., 1992; Salmerón et al., 2001; Trevisan et al., 1990). One study reported a positive association between 2-hour glucose concentrations and polyunsaturated fatty acid intake (Mooy et al., 1995). A review of epidemiological studies on this relationship concluded that higher intakes of polyunsaturated fats could be beneficial in reducing the risk for diabetes (Hu et al., 2001).

Risk of Nutrient Inadequacy

Dietary n-6 polyunsaturated fatty acids have been reported to contribute approximately 5 to 7 percent of total energy intake of adults (Allison et al., 1999; Fischer et al., 1985), and range up to no more than 10 percent of energy intake (Willett et al., 1987; Appendix Tables E-1 and E-9). The


TABLE 11-9 Interventional Studies on n-6 Fatty Acid Intake and Blood Lipid Concentrations


Study Design

Percent of Energy from Fatty Acida

Becker et al., 1983

12 men

4.3 18:2

4-wk crossover

6.8 18:2

18 18:2

Mattson and

20 adults

3.9 18:2

Grundy, 1985

4-wk crossover

29 18:2

McDonald et al.,

18 men

7.9 18:2 and 18.8 18:1


18-d parallel

21.3 18:2 and 7.0 18:1

Zock and Katan,

56 men and women

3.8 18:2 (trans diet)


3-wk crossover

3.9 18:2 (18:0 diet)

12 18:2 (18:2 diet)


30 and 33 men

7.2 ^ 1.7 18:2

et al., 1993

26-d crossover

7.2 ^ 2.1 18:2

7.2 ^ 17.8 18:2

5.7 ^ 1.6 18:2

5.7 ^ 1.8 18:2

5.7 ^ 2.1 18:2

Howard et al., 1995

63 men and women

3.0 18:2

6-wk crossover

4.2 18:2

7.0 18:2

12.8 18:2

b LDL-C = low density lipoprotein cholesterol, HDL-C = high density lipoprotein cholesterol.

main sources of n-6 polyunsaturated fatty acids are vegetable oils (e.g., soybean oil, safflower oil, and corn oil). Linoleic acid, the predominant n-6 polyunsaturated fatty acid, is essential in the diet, and therefore an Adequate Intake (AI) is set (see Chapter 8). Based on the estimated energy requirement for each age group, a minimum intake of 5 percent of energy from linoleic acid would be needed to meet the AI.


Postintervention Blood Lipid Concentration (mmol/L)'



























Quick Weight Loss Action

Quick Weight Loss Action

Why Indulge In Self-Pity When You Can Do Something About Your Weight Now. Say Goodbye to Your Weight Problems That Have Only Make Your Life Nothing But Miserable. Have you often felt short-changed because of your weight or physical appearance?

Get My Free Ebook

Post a comment