Relation between a diet with a high glycemic load and plasma concentrations of high-sensitivity C-reactive protein in middle-aged women

¹ From the Center of Cardiovascular Prevention, the Divisions of Preventive Medicine (SL, JEM, JEB, and PMR) and Cardiology (PMR) and the Channing Laboratory (JEM, MJS, and WCW), Harvard Medical School, Boston; the Department of Medicine, Brigham and Women's Hospital, and the Department of Ambulatory Care and Prevention, Harvard Medical School, Boston (JEB); and the Departments of Epidemiology (JEM, JEB, MJS, and WCW) and Nutrition (SL and WCW), the Harvard School of Public Health, Boston.

² Supported by grants PHS NO-CA47988, HL 43851, HL34595, HL58755, and DK02767 from the National Institutes of Health.

³ Reprints not available. Address correspondence to S Liu, Division of Preventive Medicine, Brigham and Women's Hospital, 900 Commonwealth Avenue East, Boston, MA 02215. E-mail: sliu{at}rics.bwh.harvard.edu.

ABSTRACT  
Background: Recent prospective data suggest that intake of rapidly digested and absorbed carbohydrates with a high dietary glycemic load is associated with an increased risk of ischemic heart disease.

Objective: We examined whether a high dietary glycemic load was associated with elevated hs-CRP concentrations and whether this association was modified by body mass index (BMI; in kg/m²).

Design: In 244 apparently healthy women, we measured plasma hs-CRP concentrations and determined average dietary glycemic loads with a validated semiquantitative food-frequency questionnaire. Using multiple regression models, we evaluated the association between dietary glycemic load and plasma hs-CRP after adjusting for age; treatment status; smoking status; BMI; physical activity level; parental history of myocardial infarction; history of hypertension, diabetes, and high cholesterol; postmenopausal hormone use; alcohol intake; and other dietary variables.

Results: We found a strong and statistically significant positive association between dietary glycemic load and plasma hs-CRP. The median hs-CRP concentration for the lowest quintile of dietary glycemic load was 1.9 mg/L and for the highest quintile was 3.7 mg/L; corresponding multivariate-adjusted geometric means were 1.4 and 3.8 mg/L, respectively (P for trend < 0.01). This association was significantly modified by BMI. Among women with a BMI 25, the multivariate-adjusted geometric mean hs-CRP concentration in the lowest quintile was 1.6 mg/L and in the highest quintile was 5.0 mg/L; however, among women with a BMI < 25, the corresponding means were 1.1 and 3.1 mg/L, respectively (P = 0.01 for interaction).

Conclusions: Dietary glycemic load is significantly and positively associated with plasma hs-CRP in healthy middle-aged women, independent of conventional risk factors for ischemic heart disease. Exacerbation of the proinflammatory process may be a mechanism whereby a high intake of rapidly digested and absorbed carbohydrates increases the risk of ischemic heart disease, especially in overweight women prone to insulin resistance.

Key Words: Dietary carbohydrate • glycemic index • glycemic load • high-sensitivity C-reactive protein • obesity • ischemic heart disease • middle-aged women • Women's Health Study

INTRODUCTION  
Inflammation plays a key role in the pathogenesis of atherothrombosis (1), and measurement of high-sensitivity C-reactive protein (hs-CRP)—a sensitive marker for systemic inflammation—can identify individuals at high risk of developing ischemic heart disease (IHD) (2). In several large prospective studies of apparently healthy men and women, elevated plasma hs-CRP concentrations were related to increased risk of IHD, independent of conventional lipid and hemodynamic risk factors for IHD (3–7). Elevated plasma hs-CRP has also been associated with obesity, insulin resistance, and hyperglycemia (8–11), suggesting that insulin resistance, type 2 diabetes, and IHD may be consequences of the ongoing acute phase response, reflecting a chronic adaptation of the immune system (12).

Diet can significantly affect insulin sensitivity and the risk of type 2 diabetes and IHD (13, 14), and the effects of specific components of the diet have been investigated (15–18). Metabolic studies have shown that a high intake of rapidly digested and absorbed carbohydrates can induce rapid postprandial glucose and insulin responses, leading to an insulin-resistant state characterized by hyperinsulinemia and dyslipidemia (ie, high triacylglycerol and low HDL concentrations) (19–22). The glycemic index was introduced to quantify and rank the physiologic responses to different carbohydrate-containing foods (23). A food's glycemic index is defined as the incremental area under the blood glucose curve induced by a specific carbohydrate-containing food and is expressed as a percentage of the area produced by the same amount of carbohydrates from a standard source, either glucose or white bread (24). Generally, foods with a high glycemic load (ie, glycemic index x carbohydrate content) cause more rapid glucose and insulin responses (20, 25–27). In the clinical management of diabetes, lowering the dietary glycemic load improves hyperglycemia and dyslipidemia (20, 21, 28–31).

We recently reported that a high dietary glycemic load is associated with low plasma HDL concentrations, elevated fasting triacylglycerol concentrations (30), and an increased risk of developing IHD (17) and type 2 diabetes (15, 32). On the basis of these data, we hypothesized that dietary glycemic load might also influence plasma concentrations of hs-CRP. In the present cross-sectional study of 244 middle-aged women in the Women's Health Study (WHS), we sought to determine whether a dietary pattern characterized by a high dietary glycemic load is associated with elevated plasma hs-CRP concentrations and whether this association is modified by body mass index (BMI; in kg/m²), as an indicator of insulin resistance.

SUBJECTS AND METHODS  
Women's Health Study subjects
The WHS is a randomized, double-blind, placebo-controlled trial designed to test the balance of benefits and risks of low-dose aspirin and vitamin E in the primary prevention of cardiovascular disease and cancer in 39876 female health professionals who had no heart disease, cancer (other than nonmelanoma skin cancer), or history of stroke at baseline. Detailed information on carbohydrate intake was provided by 39345 (98%) of the randomly assigned participants who completed a 131-item validated, semiquantitative food-frequency questionnaire (SFFQ) at baseline in 1993 (33, 34). The protocol was approved by Brigham and Women's Hospital and Harvard Medical School.

Collection of blood samples
At baseline, blood samples were collected from 28133 women (71% of the WHS participants) into tubes containing EDTA and were stored in liquid nitrogen until analyzed. For the current analysis, we included 244 randomly selected women who had served as control subjects in a recent case-control study of cardiovascular disease nested in the WHS (6). Plasma samples from these 244 women were thawed and assayed for hs-CRP with latex-enhanced immunonephelometry on a BN II analyzer (Dade Behring, Newark, DE) (35). All samples were handled identically and analyzed in random order to reduce interassay variation.

Assessment of diet with a semiquantitative food-frequency questionnaire
For each food, a commonly used unit or portion size (eg, one slice of bread) was specified on the SFFQ, and each participant was asked how often on average during the previous year she had consumed that amount. Nine responses were possible, ranging from "never" to "6 or more times per day." Nutrient scores were computed by multiplying the frequency of consumption of each unit of food from the SFFQ by the nutrient content of the specified portion size according to food-composition tables from the US Department of Agriculture (36) and other sources. A full description of the SFFQ and of the procedures used to calculate nutrient intake and data on the reproducibility and validity of the SFFQ in a similar cohort were previously reported (30, 37). The correlation coefficient for energy-adjusted carbohydrate intake between the SFFQ and diet records was 0.73 (37). The validity of the SFFQ in assessing individual foods high in carbohydrate was also documented (38).

Calculation of the dietary glycemic load
We calculated the dietary glycemic load of each food by multiplying the carbohydrate content of one serving by the glycemic index (17, 30). For example, the glycemic load of one serving of cooked potatoes was determined to be 38 because the carbohydrate content of one serving of potatoes is 37 g and the glycemic index of potatoes (with white bread as the reference) is 102% (ie, 1.02 x 37 = 38). We then multiplied this dietary glycemic load score by the frequency of consumption (1 time/d = 1, 2–3 times/d = 2.5, etc) and summed the products over all food items to produce the dietary glycemic load. The assumption made was that if a subject's usual portion was 2 servings of potato at one meal, they would be expected to double their reported frequency of intake. The dietary glycemic load thus represents the quality and quantity of carbohydrates, and each unit of dietary glycemic load is the equivalent of 1 g carbohydrate from white bread (17, 30). Additionally, the overall dietary glycemic index—a variable representing the overall quality of carbohydrate intake for each participant—was created by dividing the dietary glycemic load by the total amount of carbohydrate consumed. Representation of the dietary glycemic load per unit of carbohydrate allowed for this measure to essentially match the carbohydrate content gram by gram and thus reflects the overall quality of the carbohydrate in the entire diet.

Statistical analysis
We first calculated the means (±SDs) and proportions of covariates for this sample of women. Because plasma concentrations of hs-CRP were not distributed normally, we calculated medians and geometric means. Each major dietary variable (ie, dietary glycemic load, overall dietary glycemic index, and carbohydrate intake) was treated as either continuous or categorical (by quintiles). We then calculated the median plasma hs-CRP concentrations according to quintiles of dietary glycemic load, overall dietary glycemic index, or total carbohydrate intake. Geometric means were computed by regressing the natural logarithm of hs-CRP concentrations on dietary variables and then taking an antilog of the resulting mean logarithmic hs-CRP concentration. Multiple linear regression models were used to adjust for potential confounding factors, including age, randomized treatment status, parental history of myocardial infarction (MI) before the age of 60 y, history of diabetes mellitus, history of hypertension, history of high cholesterol, use of postmenopausal hormones, BMI, smoking status, alcohol intake, physical activity levels, and intakes of dietary fiber, folate, protein, cholesterol, and total energy. Total energy intake was adjusted by using the residual method (37).

We also calculated the differences in plasma hs-CRP concentrations corresponding to 25-unit differences in the dietary glycemic load, which correspond to 2 slices of white bread. The robust variance was used to ensure valid inference even if the regression residuals were not normally distributed (39). We also used multiple logistic regression to calculate the odds ratios of elevated hs-CRP concentrations (defined as greater than the median concentration of 2.8 mg/L in the study population) for women in the highest quintile of dietary glycemic load compared with women in the lowest quintile. Statistical analyses were conducted with the use of SAS software (version 6; SAS Institute, Cary, NC). The effect of a diet with a high glycemic load is most pronounced in persons with more severe insulin resistance, particularly if they have a high BMI (17, 30, 40, 41). We therefore repeated several of the analyses stratified by BMI (< or 25).

RESULTS  
The 244 women in the study were aged 45–82 y ( ± SD: 59 ± 8 y) and had a mean BMI of 26.0 ± 5.2 (Table 1
View this table:
TABLE 1 . Distribution of dietary and lifestyle characteristics at baseline in 244 women in the Women's Health Study¹  
As expected, plasma concentrations of hs-CRP were not normally distributed but were skewed to the right; the median hs-CRP concentration was 2.8 mg/L (interquartile range: 1.1–5.5 mg/L). The distribution of hs-CRP concentrations at baseline by covariate status is shown in Table 2. Overall, women who were obese, smoked, took postmenopausal hormones, or had a history of hypertension, high cholesterol, or diabetes mellitus tended to have higher plasma concentrations of hs-CRP.

View this table:
TABLE 2 . Distribution of high-sensitivity C-reactive protein (hs-CRP) concentrations at baseline in 244 women in the Women's Health Study by covariate status  
The median plasma concentrations of hs-CRP increased across quintiles of dietary glycemic load, overall dietary glycemic index, and total carbohydrate intake (Table 3). For increasing quintiles of dietary glycemic load, the corresponding median hs-CRP concentrations were 1.9, 2.1, 3.2, 2.8, and 3.7 mg/L. Similar relations were observed for overall dietary glycemic index and carbohydrate intake, albeit with less magnitude. In multivariate analyses, dietary glycemic load, overall dietary glycemic index, and total carbohydrate intake were each positively related to plasma hs-CRP concentrations. For the lowest and highest quintiles of dietary glycemic load, the multivariate-adjusted geometric mean hs-CRP concentrations were 1.4 and 3.8 mg/L (P < 0.01) after adjustment for all the covariates listed in Table 1. For the lowest and highest quintiles of overall dietary glycemic index, the multivariate-adjusted geometric mean hs-CRP concentrations were 1.8 and 2.8 mg/L (P < 0.01); those for the lowest and highest quintiles of total carbohydrate intake were 1.6 and 3.1 mg/L (P < 0.01). After further adjustment for carbohydrate intake, dietary glycemic load remained positively associated with hs-CRP concentrations (P < 0.001). In a multivariate model that included the same covariates but treated dietary glycemic load as a continuous variable, a 25-unit increment in dietary glycemic load was associated with a 1.5-mg/L increase in hs-CRP concentration (P = 0.03).

View this table:
TABLE 3 . Plasma high-sensitivity C-reactive protein (hs-CRP) concentrations in 244 women in the Women's Health Study by dietary glycemic load, dietary glycemic index, and carbohydrate intake  
The relation between dietary glycemic load and plasma hs-CRP concentrations differed significantly by BMI category (Figure 1). For the lowest to the highest quintiles of dietary glycemic load, the multivariate-adjusted geometric mean hs-CRP concentrations were 1.6 and 5.0 mg/L in women with a BMI 25 (n = 111) and 1.1 and 3.1 mg/L in women with a BMI < 25 (P = 0.01 for interaction).

View larger version (15K):
FIGURE 1. . Adjusted geometric mean plasma concentrations of high-sensitivity C-reactive protein (hs-CRP) by quintiles (Q1–Q5) of energy-adjusted dietary glycemic load in 244 women in 2 BMI categories: BMI < 25 ( and dashed regression line) and BMI 25 (• and solid regression line). Multiple linear regression models were used to adjust for potential confounding factors, including age, randomized treatment status, smoking status, BMI, physical activity levels, alcohol intake, parental history of myocardial infarction before the age of 60 y, history of diabetes mellitus, history of hypertension, history of high cholesterol, postmenopausal hormone use, and intakes of dietary fiber, folate, protein, cholesterol, and total energy. P = 0.01 for the interaction between BMI and dietary glycemic load. Mean dietary glycemic load for each quintile in parentheses.

 
Finally, we dichotomized hs-CRP concentrations at the median of 2.8 mg/L in the distribution and used a multiple logistic model to calculate the odds ratios of elevated hs-CRP concentrations (>2.8 mg/L) across quintiles of dietary glycemic load (Table 4). We found that the odds ratio of elevated hs-CRP concentrations for women in the highest compared with the lowest quintile of dietary glycemic load was 9.43 (95% CI: 1.92, 46.23).

View this table:
TABLE 4 . Elevated plasma high-sensitivity C-reactive protein (CRP) concentrations in 244 women in the Women's Health Study by dietary glycemic load, dietary glycemic index, and carbohydrate intake¹  

DISCUSSION  
In this sample of apparently healthy middle-aged women, both the quality and the quantity of carbohydrate intake were directly related to plasma concentrations of hs-CRP, independent of conventional risk factors for IHD. Dietary glycemic load appeared to best summarize the combined effects of the quantity and quality of the carbohydrate on plasma concentrations of hs-CRP. In particular, the adjusted geometric mean plasma hs-CRP concentration in the highest quintile of dietary glycemic load was nearly 2-fold that in the lowest quintile of dietary glycemic load. Furthermore, the dose-response gradient between dietary glycemic load and the mean plasma concentration of hs-CRP was stronger in women with a BMI 25.

Recent observational and experimental evidence indicates that inflammation plays a key role in the pathogenesis of atherothrombosis (1) and that plasma concentrations of hs-CRP can be used to identify individuals at high risk of developing IHD (2). Elevated plasma concentrations of hs-CRP were associated with an increased risk of IHD, independent of conventional lipid and hemodynamic risk factors for IHD (3–7), and with obesity, insulin resistance, and hyperglycemia (8–11). More recently, the vascular endothelium (particularly the vast capillary and arteriolar endothelium)—the same tissue involved in the development of IHD—was hypothesized to be a central site of insulin resistance (42). Taken together, these observations indicate that insulin resistance, type 2 diabetes, and IHD are partly consequences of the ongoing acute phase response to various stimuli and reflect a chronic adaptation of the immune system (12). Previous studies also showed that a high intake of low-quality carbohydrates, characterized by a high dietary glycemic index, may also lead to obesity (43) and to an increased risk of type 2 diabetes mellitus (15, 18, 32) and IHD (17, 44). Few studies, however, have directly examined the association between dietary factors and plasma concentrations of hs-CRP.

Our finding that a high dietary glycemic load predicts elevated plasma concentrations of hs-CRP has several potential implications. First, our findings suggest that dietary glycemic load may affect insulin resistance and risk of IHD through a proinflammatory process that results in increased production of hs-CRP. Recent experimental data showed that elevated concentrations of insulin and counterregulatory hormones, which are common in individuals who are obese or insulin resistant, are directly associated with hepatic production of CRP (45). Alternatively, inflammation—marked by a high hs-CRP concentration—might result from the recurrent postprandial hyperglycemia, hyperinsulinemia, and insulin resistance that may occur with a long-term high dietary glycemic load. It is also possible that hyperglycemia may lead to the production of advanced glycation end products, which may stimulate the liver to increase production of acute phase reactants. Second, our findings indicate that the dose-response gradient between dietary glycemic load and plasma hs-CRP concentrations was most apparent in overweight women. In another study, we found that a high dietary glycemic load in 280 postmenopausal nurses was associated with higher fasting triacylglycerol and lower HDL concentrations, especially in women who were overweight (30). These findings are consistent with the hypothesis that the adverse metabolic effects of a high dietary glycemic load may be magnified in persons who are obese or insulin resistant (46–48).

In the current study, women who were overweight (ie, BMI 25) and diabetic had elevated hs-CRP concentrations (Table 2). These findings are consistent with data relating obesity and type 2 diabetes to elevated plasma concentrations of hs-CRP (8–11) and with the observation that adipocytes secrete interleukin 6, a primary hepatic stimulant of CRP production.

In the Physicians' Health Study, baseline plasma CRP concentrations in initially healthy men predicted future risk of first MI and ischemic stroke (7). When compared with the men in the lowest quartile of CRP concentrations, the men in the highest quartile were 3 times as likely to have a first MI (relative risk = 2.9, P < 0.001) and 2 times as likely to have ischemic stroke (relative risk = 1.9, P = 0.02) during 9 y of follow-up. In particular, the benefits of aspirin in reducing the risk of a first MI were most evident in men in the highest quartile of CRP concentrations, suggesting an antiinflammatory mechanism. In another study conducted in the present cohort of female health professionals, each 1.0-mg/L increase in the hs-CRP concentration was associated with a 25% increase in risk of cardiovascular disease (6). Thus, the 2.4-mg/L increase in plasma hs-CRP concentrations that we observed when comparing the lowest with the highest quintile of dietary glycemic load would theoretically translate into a 60% increase in IHD risk. However, because our study was not a randomized feeding trial, we cannot exclude the possibilities of misclassification, selection bias, and residual confounding. Moreover, because there are few metabolic data on the effects of dietary modification on plasma hs-CRP concentrations, we cannot directly evaluate the consistency of our data relative to data from metabolic trials. Future randomized intervention trials are warranted to evaluate the effect of dietary determinants on plasma concentrations of hs-CRP.

Plasma hs-CRP concentrations are affected by many physiologic, infectious, and other environmental or lifestyle factors. For most individuals, however, hs-CRP concentrations appear to be stable over long periods of time. In one study that evaluated blood samples obtained 5 y apart, the correlation coefficient for hs-CRP was 0.60, a value equal to or greater than that observed for plasma measurements of lipids (49). Other factors that may affect the between-person variation of hs-CRP concentrations include the time integration of the measurements, laboratory errors, and individual metabolic differences. Nevertheless, because our SFFQ was designed to assess long-term average dietary intakes and because the errors in dietary measures assessed by the SFFQ should be independent of errors in measurements of hs-CRP, the strong association observed between dietary glycemic load and hs-CRP concentration cannot be explained by these potential independent errors. Moreover, the overall dose-response gradient of hs-CRP concentrations across quintiles of dietary glycemic load appeared to be modified by BMI. This finding was consistent with our a priori hypothesis based on results from metabolic data.

In conclusion, in the 244 apparently healthy middle-aged women in the current study, dietary glycemic load was directly related to plasma concentrations of hs-CRP, independent of BMI, total energy intake, and other known risk factors of IHD. Although these cross-sectional data do not prove cause and effect, they suggest that exacerbation of the proinflammatory process may be a mechanism whereby a high intake of rapidly digested and absorbed carbohydrates increases the risk of IHD, especially in overweight women, who are prone to insulin resistance.

REFERENCES  

1. Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 1999;340:115–26.
2. Ridker PM. Novel risk factors and markers for coronary disease. Adv Intern Med 2000;45:391–418.
3. Kuller LH, Tracy RP, Shaten J, Meilahn EN. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case-control study—Multiple Risk Factor Intervention Trial. Am J Epidemiol 1996;144:537–47.
4. Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly. Results from the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol 1997;17:1121–7.
5. Koenig W, Sund M, Frohlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation 1999; 99:237–42.
6. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342:836–43.
7. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973–9.
8. Mendall MA, Patel P, Ballam L, Strachan D, Northfield TC. C reactive protein and its relation to cardiovascular risk factors: a population based cross sectional study. BMJ 1996;312:1061–5.
9. Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care 1999;22:1971–7.
10. Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282:2131–5.
11. Hak AE, Stehouwer CD, Bots ML, et al. Associations of C-reactive protein with measures of obesity, insulin resistance, and subclinical atherosclerosis in healthy, middle-aged women. Arterioscler Thromb Vasc Biol 1999;19:1986–91.
12. Pickup JC, Mattock MB, Chusney GD, Burt D. NIDDM as a disease of the innate immune system: association of acute-phase reactants and interleukin-6 with metabolic syndrome X. Diabetologia 1997; 40:1286–92.
13. American Heart Association. Dietary guidelines for healthy American adults. Circulation 1996;94:1795–800.
14. American Diabetes Association. Nutrition recommendations and principles for people with diabetes mellitus. Diabetes Care 2000; 23(suppl):S43–6.
15. Salmeron J, Manson JE, Stampfer MJ, Colditz GA, Wing AL, Willett WC. Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA 1997;277:472–7.
16. Liu S. Dietary glycemic load, carbohydrate, and whole grain intakes in relation to risk of coronary heart disease. PhD thesis. Departments of Epidemiology and Nutrition, Harvard University, Boston, 1998.
17. Liu S, Willett WC, Stampfer MJ, et al. A prospective study of dietary glycemic load, carbohydrate intake, and risk of coronary heart disease in US women. Am J Clin Nutr 2000;71:1455–61.
18. Liu S, Manson JE, Stampfer MJ, et al. A prospective study of whole-grain intake and risk of type 2 diabetes mellitus in US women. Am J Public Health 2000;90:1409–15.
19. Jenkins DJ, Wolever TM, Collier GR, et al. Metabolic effects of a low-glycemic-index diet. Am J Clin Nutr 1987;46:968–75.
20. Wolever TM. The glycemic index. World Rev Nutr Diet 1990;62: 120–85.
21. Miller JC. Importance of glycemic index in diabetes. Am J Clin Nutr 1994;59(suppl):747S–52S.
22. Frost G, Leeds A, Trew G, Margara R, Dornhorst A. Insulin sensitivity in women at risk of coronary heart disease and the effect of a low glycemic diet. Metabolism 1998;47:1245–51.
23. Jenkins DJ, Wolever TM, Taylor RH, et al. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981;34:362–6.
24. Wolever TM, Jenkins DJ, Jenkins AL, Josse RG. The glycemic index: methodology and clinical implications. Am J Clin Nutr 1991; 54:846–54.
25. Jenkins DJA, Jenkins AL, Wolever TMS, et al. Low glycemic index: lente carbohydrates and physiological effects of altered food frequency. Am J Clin Nutr 1994;59(suppl):706S–9S.
26. Bjorck I, Granfeldt Y, Liljeberg H, Tovar J, Asp N-G. Food properties affecting the digestion and absorption of carbohydrates. Am J Clin Nutr 1994;59(suppl):699S–705S.
27. Granfeldt Y, Bjorck I, Drews A, Tovar J. An in vitro procedure based on chewing to predict metabolic response to starch in cereal and legume products. Eur J Clin Nutr 1992;46:649–60.
28. Jenkins D, Wolever T, Kalmusky J, et al. Low-glycemic index diet in hyperlipidemia: use of traditional starchy foods. Am J Clin Nutr 1987;46:66–71.
29. Frost G, Leeds A, Dore D, Madeiros S, Brading S, Dornhorst A. Glycemic index as a determinant of serum HDL-cholesterol concentration. Lancet 1999;353:1045–8.
30. Liu S, Manson JE, Stampfer MJ, et al. Dietary glycemic load assessed by food-frequency questionnaire in relation to plasma high-density-lipoprotein cholesterol and fasting triacylglycerols in postmenopausal women. Am J Clin Nutr 2001;73:560–6.
31. Ford E, Liu S. Glycemic index and serum high-density lipoprotein (HDL) cholesterol concentration among United States adults. Arch Intern Med 2001;161:572–6.
32. Salmeron J, Ascherio A, Rimm EB, et al. Dietary fiber, glycemic load, and risk of NIDDM in men. Diabetes Care 1997;20:545–50.
33. Liu S, Manson JE, Lee IM, et al. Fruit and vegetable intake and risk of cardiovascular disease: the Women's Health Study. Am J Clin Nutr 2000;72:922–8.
34. Willett WC. Nutritional epidemiology. In: Rothman KJ, Greenland S, eds. Modern epidemiology. 2nd ed. Philadelphia: Lippincott-Raven Publishers, 1998.
35. Roberts WL, Sedrick R, Moulton L, Spencer A, Rifai N. Evaluation of four automated high-sensitivity C-reactive protein methods: implications for clinical and epidemiological applications. Clin Chem 2000;46:461–8.
36. US Department of Agriculture. Composition of foods—raw, processed, and prepared, 1963–1988. Agricultural handbook no. 8. Washington, DC: US Government Printing Office, 1989.
37. Willett WC. Nutritional epidemiology. 2nd ed. New York: Oxford University Press, 1998.
38. Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol 1989;18:858–67.
39. Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986;42:121–30.
40. Jeppesen J, Schaaf P, Jones C, Zhou MY, Chen YD, Reaven GM. Effects of low-fat, high-carbohydrate diets on risk factors for ischemic heart disease in postmenopausal women. Am J Clin Nutr 1997;65:1027–33.
41. Liu S. Insulin resistance, hyperglycemia and risk of major chronic diseases—a dietary perspective. J Aust Nutr Soc 1998;22:140–50.
42. Pinkney JH, Stehouwer CD, Coppack SW, Yudkin JS. Endothelial dysfunction: cause of the insulin resistance syndrome. Diabetes 1997;46(suppl):S9–13.
43. Ludwig DS, Majzoub JA, Al-Zahrani A, Dallal GE, Blanco I, Roberts SB. High glycemic index foods, overeating, and obesity. Pediatrics [serial online] 1999;103:e26. Internet: http://www.pediatrics.org (accessed 14 December 2001).
44. Liu S, Willett W, Hu F, et al. Dietary glycemic load, carbohydrates and risk of coronary heart disease among diabetic women. Circulation 1998;98(suppl):I–193 (abstr).
45. O'Riordain MG, Ross JA, Fearon KC, et al. Insulin and counterregulatory hormones influence acute-phase protein production in human hepatocytes. Am J Physiol 1995;269:E323–30.
46. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988;37:1595–607.
47. Mensink RP, Katan MB. Effect of dietary fatty acids on serum lipids and lipoproteins: a meta-analysis of 27 trials. Arterioscler Thromb 1992;12:911–9.
48. Knopp RH, Walden CE, Retzlaff BM, et al. Long-term cholesterol-lowering effects of 4 fat-restricted diets in hypercholesterolemic and combined hyperlipidemic men. The Dietary Alternatives Study. JAMA 1997;278:1509–15.
49. Ridker PM, Rifai N, Pfeffer MA, Sacks F, Braunwald E. Long-term effects of pravastatin on plasma concentration of C-reactive protein. The Cholesterol and Recurrent Events (CARE) Investigators. Circulation 1999;100:230–5.