AJCN EB Program 2010
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.
Agricola
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.
American Journal of Clinical Nutrition, Vol. 87, No. 6, 1793-1801, June 2008
© 2008 American Society for Nutrition

ORIGINAL RESEARCH COMMUNICATION

Glycemic index, glycemic load, and cancer risk: a meta-analysis1,2,3

Patrizia Gnagnarella, Sara Gandini, Carlo La Vecchia and Patrick Maisonneuve

1 From the Division of Epidemiology and Biostatistics, European Institute of Oncology, Milan, Italy (PG, SG, and PM); the Istituto di Ricerche Farmacologiche "Mario Negri," Milan, Italy (CLV); and the Istituto di Statistica Medica e Biometria "Giulio A Maccacaro," Unità di Statistica Medica, Università degli Studi di Milano, Milan, Italy (CLV)

2 Supported by grants from the Italian Association for Cancer Research (AIRC) and by a Senior Fellowship of the International Agency for Research on Cancer (CLV).

3 Reprints not available. Address correspondence to P Gnagnarella, Division of Epidemiology and Biostatistics, European Institute of Oncology, Via Ripamonti 435, 20141 Milan, Italy. E-mail: patrizia.gnagnarella{at}ieo.it.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: Factors linked to glucose metabolism play an important role in the development of cancers, and both glycemic index (GI) and glycemic load (GL) have been investigated as potential etiologic factors.

Objective: A meta-analysis was performed to explore the association between GI and GL and cancer risk from published studies.

Design: A comprehensive, systematic bibliographic search of the medical literature was conducted to identify relevant studies. Case-control and cohort studies published before October 2007 that reported cancer risk estimates for GI and GL were included. Pooled relative risks (RRs) were estimated for breast, colorectal, endometrial, and pancreatic cancer.

Results: Thirty-nine studies were included in the meta-analysis. The interquantile ranges of GL were significantly wider in case-control studies, most of which were conducted in European countries, than in cohort studies. Cohort studies that presented lower ranges of GL also reported lower risk estimates. Overall, both GL and GI were significantly associated with a greater risk of colorectal (summary RR = 1.26; 95% CI: 1.11, 1.44 and RR = 1.18; 95% CI: 1.05, 1.34, respectively) and endometrial (RR = 1.36; 95% CI: 1.14, 1.62 and RR = 1.22; 95% CI: 1.01, 1.49) cancer than of breast and pancreatic cancer. There was, however, a significant between-study heterogeneity for colorectal cancer (P < 0.0001). The association between GL and breast cancer disappeared when publication bias was taken into account. No association was found for pancreatic cancer.

Conclusion: This comprehensive meta-analysis of GI and GL and cancer risk suggested an overall direct association with colorectal and endometrial cancer.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Glycemic index (GI) is a ranking system for carbohydrates according to their effect on blood glucose concentrations. It compares available carbohydrates gram-for-gram in individual foods, providing a numerical, evidence-based index of postprandial glycemia (1, 2). The glycemic load (GL) combines the qualitative and quantitative measures of carbohydrate and consists of the sum of the GLs for the total servings of all carbohydrate-containing foods consumed per day, on average. Therefore, the overall GI reflects the average quality of carbohydrates consumed, whereas the total dietary GL reflects both the average quantity and quality of carbohydrates (3).

Numerous studies have investigated GI and GL as potential risk factors for several chronic diseases. In fact, factors linked to glucose metabolism seem to play an important role in the development of cardiovascular disease, diabetes, obesity, and cancer (4–6). In particular, high insulin concentrations have been suggested as a potential unifying mechanism for the risk of several types of cancer (4, 7). The main mechanism by which a high-GI diet may increase cancer risk is modulation of the insulin-like growth factor (IGF) axis. Insulin acts as a growth factor for colonic mucosal cells per se, and it can increase the activity of insulin-like growth factors, such as IGF-I, which in turn stimulate cell proliferation and differentiation and can inhibit apoptosis (8). Insulin also can suppress hepatic secretion of IGF-binding protein-1, influence sex hormone concentrations, and reduce the concentrations of their binding proteins (9, 10). Other conditions such as insulin resistance, hyperglycemia, obesity, or diabetes also can influence cancer risk(4, 5, 7, 8, 11). In view of the growing number of published reports on the relation between GI and GL and cancer risk and of the apparent inconsistency in the results, we decided to conduct a meta-analysis to better quantify the magnitude of the risk and to identify potential sources of variability.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Data sources and search strategy
Published reports were identified from electronic databases [ie, PubMed, ISI Web of Science (Science Citation Index Expanded), and Embase] by using validated search strategies (12–14). Additional reports were extracted from the reference lists of the retrieved articles and reviews on the topic. The key words used for the literature search included cancer, malignancy, glycemic index, glycemic load, and diet. We also retrieved all case-control and cohort studies that investigated the association between dietary carbohydrate and sugar intake and cancer at any site. The search was limited to studies in humans, but no language or time restrictions were applied.

We included any case-control, cohort, or cross-sectional study that was published as an original article before October 2007 and that reported sufficient information to allow an adequate estimation, at least for the relative risk (RR) for the upper versus lowest quantile (and 95% CIs). That is, the study reported either adjusted odds ratios or RRs or crude data and SEs, variance, CIs, or P values of the significance of the estimates. We excluded multiple reports based on the same study population.

Selection of articles
Two investigators independently reviewed all of the studies, and data extraction was cross-checked to improve accuracy. Three reports were excluded from the meta-analysis: 2 studies were not independent (15, 16), and 1 provided dose-response risk estimates but no information on the range of exposure (17). For one study (18), we obtained the missing information directly from the authors. The study of Oh et al (19) included estimates only for colorectal cancer and thus was excluded from this analysis; it was included in a subanalysis by Michaud et al (20) of the same cohort (Nurses’ Health Study).

For each study included in the meta-analysis, we retrieved information on several factors: study—publication year, study design, and study location; cases and controls—number and source; population—premenopausal or postmenopausal status for breast and endometrial cancer; subsites—colon or rectum; exposure—dietary assessment methods used, origin of the glycemic values assigned to foods, methods applied to calculate GI and GL, values used to define the upper category (Qmax: the lowest value of the highest category) and the lowest category taken as reference (Qmin: the upper value of the reference category), and the interquantile range (Qmax–Qmin); and statistics—statistical methods used and adjustment for confounding variables.

Statistical analysis
We ignored the distinction between the various measures of RR (ie, odds ratio, rate ratio, and risk ratio). Each measure of association and its corresponding confidence limits were transformed into log RR, and the corresponding variance was calculated by using Greenland's formula (21). When they were available, we used estimates adjusted for the maximum number of confounding variables. When they were not given, we calculated risk estimates from tabular data and evaluated the SE of the log odds ratio by using Woolf's formula. When they were available, we used stratified risk estimates for sex, subsite, or menopausal status for women. Two-sided Wilcoxon's tests were used to compare the values of the lower and upper quantiles and interquantile ranges by study design, geographic location, and food sources used for the calculation of GL.

We assessed the homogeneity of the effect across studies by using the large-sample test based on the chi-square statistic. Because the chi-square test has limited power, we considered heterogeneity to be significant at P = 0.10 (22). Heterogeneity across studies was evaluated by I2, which represents the percentage of total variation across studies that is attributable to heterogeneity rather than to chance. We estimated summarized RRs by pooling the study-specific estimates with the use of classic random-effects models. When several measures of RR were given for a single study, random-effects models were used, including the 2 sources of variation (within and between studies), to take into account correlation within the study, by using PROC MIXED in SAS software [version 8.02; SAS Institute, Cary, NC (23, 24)]. We carried out subgroup analyses and meta-regression analyses with random-effects models to investigate heterogeneity. To maximize the power of the study, we analyzed the data for all cancer sites in a single model. However, we presented data only for the cancer sites for which the pooled RR was derived from >2 studies.

The estimates, obtained from the least-square means of the random-effects model, were adjusted for significant factors explaining heterogeneity. As sensitivity analysis, we also presented the pooled RRs obtained from stratified analyses based on cancer-specific studies.

For dose-response estimates, we retrieved the RRs and CIs and the number of cases, controls, or subjects at risk by each category of exposure. Within each study, we used a linear model to estimate the RRs associated with an increase in GL or GI of 1 unit/d. We assigned to each category of GL or GI intake the value corresponding to the midpoint of the range. Whenever possible, we obtained the summary RR by pooling the study-specific estimates by the random-effects models proposed by Greenland and Longnecker, which adjust the estimates for within-study covariance and accounts for the correlation between estimates (25). Two funnel plot–based approaches were used for assessing publication bias: the Copas and Shi method and the funnel plot regression of ln(RR) on the sample size, weighted by the inverse of the pooled variance (26, 27).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Thirty-nine published reports were evaluated in this meta-analysis. In 36 studies, dietary intake was assessed with the use of a food-frequency questionnaire (FFQ); the remaining 3 studies used history methods (28–31). All studies estimated usual dietary intake by converting the reported frequency of consumption of food and the food's portion size into estimates of nutrient intake, with the use of national food-composition tables. All authors reported or quoted the reproducibility or the validity (or both) of the FFQs.

Glycemic index and glycemic load data
Because none of the national food-composition tables include information on GI and GL, their calculations were based on data from various sources. In 14 articles (36%), calculations were based on data presented in the first edition of the international tables by Foster-Powell and Brand-Miller (32); in 19 articles (49%), calculations were based on data in revised tables published in 2002 (3). In those studies, further sources were also used: published sources (33–35; also F Brighenti, unpublished observations, 2007) or data from internal laboratories (20, 36). In the remaining 6 studies (15%), Murtaugh et al (29), De Stefani et al (37), and Slattery et al (28) reported the use of data published by Jenkins et al (1, 2, 38, 39), whereas McCarl et al (40), Folsom et al (41), and Johnson et al (42) used data published by Wolever et al (43, 44).

Either white bread or glucose can be used as standard food to determinate the GI (the GI value obtained by using white bread as standard food is {approx}1.4 times that obtained by using glucose) (1, 45, 46). Most (n = 25; 64%) of the studies used white bread as reference food, and 6 (15%) used glucose; in the remaining 8 (21%) studies, the reference used was not clearly stated.

Meta-analysis
Overall, 39 reports (24 from cohort studies and 15 from case-control studies) with data on cancer and GL or GI were included in this meta-analysis (Table 1Go). Eleven of the case-control studies were hospital-based, and all 11 studies excluded from the control group subjects with diseases of the digestive tract or in any case associated with long-term modifications of the diet (eg, diabetes mellitus, ulcerative colitis, and Crohn disease). Cohort studies vary by length of follow-up from 5 (73) to 20 (37) y; 6 of 24 cohort studies presented a follow-up of <10 y. Most (n = 18) of those 24 cohort studies assessed the diet only at baseline. Most authors describing cohort studies do not report data on loss of follow-up, and, when they are provided, those data are very low. The data are based on record linkage to cancer registries or national databases. Eight reports from cohort studies presented RRs adjusted for diabetes. All of the estimates included in the meta-analysis, except those from Johnson et al (42), are adjusted for total energy intake or alcohol consumption or noncarbohydrate energy intake.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Description of studies included in the meta-analysis1

 
We investigated the range between the highest and the lowest category of GL in the various reports according to study features: geographic location, study design, and food source (Table 2Go). The upper quantile (Qmax) of GL was significantly higher in case-control studies, in European studies (borderline significant), and in studies using white bread as a reference food source. The interquantile ranges were significantly wider in case-control studies (which were mainly European), because of the extreme higher and lower values. For GI, we did not find significant differences by study features. The median and interquantile range were 81 (range: 78–83) for Qmax, 69 (63–71) for the lower quantile (Qmin), and 9 (8–12) for the interquantile range.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Reference category, upper category, and glycemic load values by study design, area, and reference food1

 
In Figure 1Go, the magnitude of the RR for the highest versus the lowest quantile of GL intake is plotted according to the interquantile range of GL. Cohort studies that presented lower ranges of GL intake than did case-control studies also had lower RRs.


Figure 1
View larger version (13K):
[in this window]
[in a new window]

  		FIGURE 1.. Relative risks (RRs) of any cancer for the highest versus lowest quantile of glycemic load, by interquantile range of glycemic load intake in the study population and by study design.

 
Forest plots of the RR for the highest versus lowest quantile of GL or GI intake for breast, colorectal, endometrial, and pancreatic cancers are presented in Figure 2Go and Figure 3Go. For breast and colorectal cancer, for which more reports are available, we also presented the pooled RRs by study design. For the other cancer sites, we presented also the estimates obtained from the stratified analyses, with each cancer site evaluated on its own.


Figure 2
View larger version (23K):
[in this window]
[in a new window]

  		FIGURE 2.. A: Forest plots for cancer by study design. CO, cohort studies; CC, case-control studies; Pre, premenopausal; Post, postmenopausal; M, male; F, female; C, colon; R, rectum. A: Breast cancer B: Colorectal cancer.

 

Figure 3
View larger version (31K):
[in this window]
[in a new window]

  		FIGURE 3.. Forest plots for endometrial and pancreatic cancer. M, male; F, female.

 
The pooled RRs for breast, colorectal, endometrial, and pancreatic cancer for GL and GI are shown in Table 3Go, together with the P value from chi-square tests and I2 for heterogeneity among the estimates. P values from the meta-regression models for the factors that explained between-study heterogeneity are shown in Table 4Go. Pooled estimates were adjusted for study design but not for interquantile range, even if both were significant for GL, because interquantile range and Qmax were not independent from study design. We investigated the variability among the estimates by looking also at menopausal status for endometrial and breast cancer, the subsite for colorectal cancer, sex, and the origin of the glycemic values assigned to food, but these factors did not explain between-study heterogeneity. The only factors explaining heterogeneity are indicated in Table 3Go with their P values.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Values for selected cancers by highest versus lowest intake of glycemic load and glycemic intake1

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. Between-study heterogeneity for highest and lowest intake of glycemic load and glycemic index1

 
We also carried out heterogeneity analyses in subgroups of studies defined by study design. Among cohort studies, the length of follow-up and the number of assessments of diet did not explain between-study variability. Considering features of case-controls, we observed that population-based studies presented significantly (P = 0.006) lower estimates for GI than did hospital-based studies [RR = 1.05 (95%CI: 0.86, 1.29) and RR = 1.61 (95%CI: 1.33, 1.94) for population- and hospital-based studies, respectively]. We observed GI ranges that were wider (P = 0.08) in population-based case-control studies (median: 13; interquartile range: 12–13) than in hospital-based studies (median: 9; interquartile range: 9–10). The source of the controls did not explain between-study heterogeneity for GL. Statistically significant elevated risks for the highest versus lowest quantile of GL and GI were observed for colorectal and endometrial cancer: the adjusted pooled RRs were 1.26 (95% CI: 1.11, 1.44) and 1.18 (1.05, 1.34) for colorectal cancer and 1.36 (1.14, 1.62) and 1.22 (1.01, 1.49) for endometrial cancer.

For the colorectal and endometrial cancer sites, we also evaluated dose risk. For colorectal cancer, the pooled RRs associated with a GL or GI increase of 1 unit/d did not show a significant increase—1.0004 (0.9991, 1.0016) and 1.002 (0.999, 1.005) for GL and GI, respectively—which corresponds to a pooled RR with a GL or GI increase of 10 unit/d, 1.00 (0.99, 1.02) and 1.02 (0.99, 1.05), respectively. For endometrial cancer, we observed a moderate but statistically significant increase in risk for GL—1.0008 (1.0000, 1.0016)—but not for GI—1.001 (0.999, 1.004)—which corresponds to a GL or GI increase of 10 units/d, 1.01 (1.00, 1.02) and 1.01 (0.99, 1.04), respectively.

Publication bias
We found no evidence of publication bias for colorectal or endometrial cancer by using either the Copas and Shi tests or the regression method, but we found an indication of publication bias for breast cancer for GL (P = 0.002, Copas and Shi tests; P = 0.08, regression method). When we adjusted the breast cancer estimates for publication bias, by adding 7 more studies, the indication of publication bias disappeared (P = 0.14), but the pooled RR lost significance (0.96; 95% CI: 0.85, 1.06).

Other sites
For stomach cancer, only 2 studies were published. An Italian case-control study (47) reported a significant association with GL (RR = 1.94; 95% CI: 1.47, 2.55), whereas the Swedish Mammography Cohort Study (48) did not provide evidence of a positive association. For ovarian cancer, the 2 available studies, an Italian case-control study (73) and a Canadian cohort study (49), reported a significant effect for GL (1.65; 95% CI: 1.30, 2.09 and 1.72; 95% CI: 1.13, 2.62, respectively). Only the Italian case-control study found a significant effect for GI (1.65; 95% CI: 1.30, 2.09). Another Italian case-control study indicated an effect for both GL and GI for prostate cancer (1.41; 95% CI: 1.04, 1.89 and 1.57; 95% CI: 1.19, 2.07, respectively), upper digestive tract neoplasms (1.8; 95% CI: 1.1, 2.9 and 1.5; 95% CI: 1.1, 2), and thyroid cancer (2.17; 95% CI: 1.50, 3.15 and 1.73; 95% CI: 1.20, 2.50). An Uruguayan case-control study (37) found a significant RR of lung cancer for GI (2.77; 95% CI: 1.28, 5.97).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
This comprehensive meta-analysis of GI and GL and cancer risk suggested an overall direct association for colorectal and endometrial cancer and an inconsistent relation with several other cancers. There was an indication of a modestly greater risk of breast cancer with GL than with GI, which disappeared when publication bias was taken into account.

In case-control studies, the upper quantile of GL was significantly greater than that in cohort studies, whereas the lowest quantiles did not differ significantly. Therefore, the interquantile ranges of GL were significantly wider in case-control than in cohort studies. This suggests that case-control studies included subjects with higher carbohydrate intake, because total dietary GL reflects both the average quantity and the quality of carbohydrates consumed. These case-control studies tended to capture a wider range of total dietary GL through the consumption, both broad and in larger quantities, of carbohydrates with a range of low and high GIs. Moreover, populations studied in cohorts seem to consume a more homogeneous diet with a narrower range of GL (median = 46), which probably made the detection of an association difficult, because neither GL or GI appear to represent strong determinants of disease risk (74). The different results in cohort and case-control studies also may reflect the presence of greater bias and confounding in case-control studies than in cohort studies, if disease status or other covariates affect the report of type and quantity of carbohydrates in cancer cases. The study design of case-control studies also could affect the results; in fact, we found that hospital-based studies presented significantly greater RR estimates for GI intake than did population-based case-control studies, and these differences were not explained by GI ranges that were wider in the population-based studies. In fact, hospitalized subjects may have dietary habits different from those of the general population, although all hospital-based studies excluded subjects admitted for conditions associated with dietary modifications.

In contrast, very few cohort studies assessed changes in dietary intake during follow-up. However, the populations studied, rather than the study design or the dietary instrument used, are probably responsible for these results. In fact, case-control studies were mainly conducted in European populations (67%), who tend to have a high consumption of refined carbohydrates, bread, potatoes, pasta, and rice (75–77). High carbohydrate intake may be an indicator of a high endogenous insulin environment, and insulin per se and the modulation of IGF-1 have been shown to act as cancer-promoting agents for several types of cancer. In contrast, most cohort studies (73%) were conducted in the United States and Canada, whose populations tend to consume a diet that is richer in fat and fatty foods and that hence has lower GL concentrations than the diets of most other countries.

The pooled estimates obtained in the stratified analysis for several cancer sites were not statistically significant. The inconsistencies with the results obtained with the model based on all studies are mainly due to the adjustment for study design, which explained most of the variability and was set as a fixed factor for all cancer sites.

We found a significant association between dietary GI and GL and endometrial cancer risk with no major sign of heterogeneity. The association was consistently found for GL in all of the reports and some studies found positive associations in selected subgroups, such as women with a high body mass index (in kg/m2) (41, 50), overweight women with low physical activity (51, 52), women with diabetes (41), and postmenopausal women using hormone replacement therapy (51). Obesity is a major determinant of insulin resistance and hyperinsulinemia (11, 78), and physical activity strongly influences glucose tolerance and insulin sensitivity (78, 79); thus, it is likely that persons who are overweight or inactive have a greater insulin response to their diet than do lean or active persons.

We found significant between-study heterogeneity for colorectal and breast cancer. For colorectal cancer, 3 case-control studies (28, 53, 54) reported a significantly greater risk associated with a diet high in GL and GI, whereas only 1 case-control study, conducted in California (29), reported no association. Most of the cohort studies found no significant association overall, and an association between high GI and GL and colorectal cancer was reported only in obese women (40). In the Health Professionals Follow-up Study (20), an elevated risk of colorectal cancer was observed among men with high intakes of GL, fructose, and sucrose. In the Breast Cancer Detection Demonstration Project study, Strayer et al (55) found a nonsignificant reduction of colorectal cancer risk for subjects with a diet high in carbohydrates and GI. Weijenberg et al (56) reported a significantly lower colon cancer risk for men in the highest GI quintile, especially with respect to the distal colon, but a greater risk for rectal cancer. In women, they observed a positive association between GL and proximal colon cancer risk, which was attenuated after the exclusion of the first 2 y of follow-up; this finding indicated that preclinical disease may have affected dietary intakes. These mixed results may be due to other confounding factors. The etiology of colorectal cancer is poorly understood, and it is still not clear which dietary factors could be important. In relation to carbohydrates, available data suggest that high intakes of dietary fiber may reduce the risk for colorectal cancer, and high intakes of sucrose may increase the risk (80). The consumption of dietary fiber or certain types of carbohydrates (eg, sucrose, starch, and fructose) may influence results in different directions.

For breast cancer, only 2 Italian studies, a case-control study (57) and a prospective study (58), found a positive association for both GI and GL. One study found a positive association only for GL (60), whereas other studies found a positive association for both GI and GL in the selected subgroups of postmenopausal women (59, 60) or premenopausal women with low level of physical activity (62). McCann et al (63) reported an inverse association in postmenopausal women with a high body mass index. The reasons underlying these results are unclear, although other authors (81) addressed the issue of a correspondence with an inverse association between hemoglobin A1c (HbA1c) concentrations (a marker of glycemic control) and insulin concentrations in women with postmenopausal breast cancer.

The absence of a clear association between GI or GL and cancers of the breast and pancreas does not disprove the IGF axis hypothesis for these neoplasms (82), but it probably underlines the complexity of IGF signaling and the concentrations of IGF-binding protein (83, 84). GI or GL (or both) also may influence the concentrations and composition of serum lipids, C-reactive protein (CRP), and other markers of inflammation, but available data are still open to discussion (5, 85). Furthermore, the influence of modifications in these factors on the risk of cancer, including breast, endometrial, and colorectal cancer, remains undefined (80, 86, 87).

In addition, the lack of association with GL or GI may be due in part to an inability of the dietary assessment method used in the present study—mainly FFQs—to capture the true range of GI and GL. None of these methods were specifically developed to assess GI and GL, and the limited variety of food items listed in the FFQ can limit the detectable range of GI and GL. This limitation can lead to some degree of misclassification of exposure and to an underestimation of any relation with cancer risk.

In conclusion, we found evidence of an association between high consumption of GL and GI and the risk of colorectal or endometrial cancer. The magnitude of the risk for subjects categorized into the highest versus lowest quantile of GL or GI intake is modest, however, and it varies according to study populations.


   ACKNOWLEDGMENTS
 
We thank Anja Olsen for sending supplementary data from her study.

The authors' responsibilities were as follows—PG and SG: study design, data collection, and manuscript writing; SG: statistical analyses; and PM and CLV: revision of the manuscript and significant interpretation of the results. None of the authors had a personal or financial conflict of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. 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.[Abstract/Free Full Text]
  2. Jenkins DJ, Wolever TM, Jenkins AL, Josse RG, Wong GS. The glycaemic response to carbohydrate foods. Lancet 1984;2:388–91.[Medline]
  3. Foster-Powell K, Holt SH, Brand-Miller JC. International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 2002;76:5–56.[Abstract/Free Full Text]
  4. Augustin LS, Franceschi S, Jenkins DJ, Kendall CW, La Vecchia C. Glycemic index in chronic disease: a review. Eur J Clin Nutr 2002;56:1049–71.[Medline]
  5. Ludwig DS. The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease. JAMA 2002;287:2414–23.[Abstract/Free Full Text]
  6. Brand-Miller JC. Glycemic load and chronic disease. Nutr Rev 2003;61:S49–55.[Medline]
  7. Jenkins DJ, Kendall CW, Augustin LS, et al. Glycemic index: overview of implications in health and disease. Am J Clin Nutr 2002;76(suppl):266S–73S.[Abstract/Free Full Text]
  8. Kaaks R, Lukanova A. Energy balance and cancer: the role of insulin and insulin-like growth factor-I. Proc Nutr Soc 2001;60:91–106.[Medline]
  9. Poretsky L, Kalin MF. The gonadotropic function of insulin. Endocr Rev 1987;8:132–41.[Abstract/Free Full Text]
  10. Garzo VG, Dorrington JH. Aromatase activity in human granulosa cells during follicular development and the modulation by follicle-stimulating hormone and insulin. Am J Obstet Gynecol 1984;148:657–62.[Medline]
  11. Kaaks R, Lukanova A, Kurzer MS. Obesity, endogenous hormones, and endometrial cancer risk: a synthetic review. Cancer Epidemiol Biomarkers Prev 2002;11:1531–43.[Abstract/Free Full Text]
  12. Haynes RB, Wilczynski N, McKibbon KA, Walker CJ, Sinclair JC. Developing optimal search strategies for detecting clinically sound studies in MEDLINE. J Am Med Inform Assoc 1994;1:447–58.[Abstract/Free Full Text]
  13. Shojania KG, Bero LA. Taking advantage of the explosion of systematic reviews: an efficient MEDLINE search strategy. Eff Clin Pract 2001;4:157–62.[Medline]
  14. Dickersin K, Scherer R, Lefebvre C. Identifying relevant studies for systematic reviews. BMJ 1994;309:1286–91.[Abstract/Free Full Text]
  15. Holmes MD, Liu S, Hankinson SE, Colditz GA, Hunter DJ, Willett WC. Dietary carbohydrates, fiber, and breast cancer risk. Am J Epidemiol 2004;159:732–9.[Abstract/Free Full Text]
  16. Frazier AL, Li L, Cho E, Willett WC, Colditz GA. Adolescent diet and risk of breast cancer. Cancer Causes Control 2004;15:73–82.[Medline]
  17. Giles GG, Simpson JA, English DR, et al. Dietary carbohydrate, fibre, glycaemic index, glycaemic load and the risk of postmenopausal breast cancer. Int J Cancer 2006;118:1843–7.[Medline]
  18. Nielsen TG, Olsen A, Christensen J, Overvad K, Tjonneland A. Dietary carbohydrate intake is not associated with the breast cancer incidence rate ratio in postmenopausal Danish women. J Nutr 2005;135:124–8.[Abstract/Free Full Text]
  19. Oh K, Willett WC, Fuchs CS, Giovannucci EL. Glycemic index, glycemic load, and carbohydrate intake in relation to risk of distal colorectal adenoma in women. Cancer Epidemiol Biomarkers Prev 2004;13:1192–8.[Abstract/Free Full Text]
  20. Michaud DS, Fuchs CS, Liu S, Willett WC, Colditz GA, Giovannucci E. Dietary glycemic load, carbohydrate, sugar, and colorectal cancer risk in men and women. Cancer Epidemiol Biomarkers Prev 2005;14:138–47.[Abstract/Free Full Text]
  21. Greenland S. Quantitative methods in the review of epidemiologic literature. Epidemiol Rev 1987;9:1–30.[Free Full Text]
  22. Sterne JA, Juni P, Schulz KF, Altman DG, Bartlett C, Egger M. Statistical methods for assessing the influence of study characteristics on treatment effects in ‘meta-epidemiological’ research. Stat Med 2002;21:1513–24.[Medline]
  23. van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21:589–624.[Medline]
  24. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.[Medline]
  25. Greenland S, Longnecker MP. Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. Am J Epidemiol 1992;135:1301–9.[Abstract/Free Full Text]
  26. Copas JB, Shi JQ. A sensitivity analysis for publication bias in systematic reviews. Stat Methods Med Res 2001;10:251–65.[Abstract/Free Full Text]
  27. Macaskill P, Walter SD, Irwig L. A comparison of methods to detect publication bias in meta-analysis. Stat Med 2001;20:641–54.[Medline]
  28. Slattery ML, Benson J, Berry TD, et al. Dietary sugar and colon cancer. Cancer Epidemiol Biomarkers Prev 1997;6:677–85.[Abstract/Free Full Text]
  29. Murtaugh MA, Sweeney C, Ma KN, et al. Vitamin D receptor gene polymorphisms, dietary promotion of insulin resistance, and colon and rectal cancer. Nutr Cancer 2006;55:35–43.[Medline]
  30. Cust AE, Slimani N, Kaaks R, et al. Dietary carbohydrates, glycemic index, glycemic load, and endometrial cancer risk within the European Prospective Investigation into Cancer and Nutrition cohort. Am J Epidemiol 2007;166:912–23.[Abstract/Free Full Text]
  31. Margetts BM, Pietinen P. European Prospective Investigation into Cancer and Nutrition: validity studies on dietary assessment methods. Int J Epidemiol 1997;26(suppl):S1–5.[Free Full Text]
  32. Foster-Powell K, Miller JB. International tables of glycemic index. Am J Clin Nutr 1995;62(suppl):871S–90S.[Medline]
  33. Brighenti F, Casiraghi M. Influenza dei processi di trasformazione sulla risposta glicemica ad alimenti amidacei. (The influence of transformation processes on the glycemic response to starchy foods.) Giornale Italiano di Nutrizione Clinica e Preventiva 1992;1:79 (in Italian).
  34. Henry CJ, Lightowler HJ, Strik CM, Renton H, Hails S. Glycaemic index and glycaemic load values of commercially available products in the UK. Br J Nutr 2005;94:922–30.[Medline]
  35. Human Nutrition Unit, University of Sidney. The official website of the glycemic index and GI database. Internet: http://www.glycemicindex.com/ (accessed 11 October 2007).
  36. Michaud DS, Liu S, Giovannucci E, Willett WC, Colditz GA, Fuchs CS. Dietary sugar, glycemic load, and pancreatic cancer risk in a prospective study. J Natl Cancer Inst 2002;94:1293–300.[Abstract/Free Full Text]
  37. De Stefani E, Deneo-Pellegrini H, Mendilaharsu M, Ronco A, Carzoglio JC. Dietary sugar and lung cancer: a case-control study in Uruguay. Nutr Cancer 1998;31:132–7.[Medline]
  38. Jenkins DJ. Lente carbohydrate: a newer approach to the dietary management of diabetes. Diabetes Care 1982;5:634–41.[Abstract]
  39. Jenkins DJ, Wolever TM, Kalmusky J, et al. Low glycemic index carbohydrate foods in the management of hyperlipidemia. Am J Clin Nutr 1985;42:604–17.[Abstract/Free Full Text]
  40. McCarl M, Harnack L, Limburg PJ, Anderson KE, Folsom AR. Incidence of colorectal cancer in relation to glycemic index and load in a cohort of women. Cancer Epidemiol Biomarkers Prev 2006;15:892–6.[Abstract/Free Full Text]
  41. Folsom AR, Demissie Z, Harnack L. Glycemic index, glycemic load, and incidence of endometrial cancer: the Iowa women's health study. Nutr Cancer 2003;46:119–24.[Medline]
  42. Johnson KJ, Anderson KE, Harnack L, Hong CP, Folsom AR. No association between dietary glycemic index or load and pancreatic cancer incidence in postmenopausal women. Cancer Epidemiol Biomarkers Prev 2005;14:1574–5.[Free Full Text]
  43. Wolever TM, Vuksan V, Katzman-RL, et al. Glycaemic index of some fruits and fruits products in patients with diabetes. Int J Food Sci Nutr 1993;43:205–12.
  44. Wolever TM, Katzman-RL, Jenkins AL, Vuksan V, Josse RG, Jenkins DJ. Glycaemic index of 102 complex carbohydrate foods in patients with diabetes. Nutr Res 1994;14:651–69.
  45. FAO. Carbohydrates in human nutrition (FAO Food and Nutrition paper no. 66). Rome, Italy: Food Agriculture Organization, 1998;66.
  46. Wolever TM, Nuttall FQ, Lee R, et al. Prediction of the relative blood glucose response of mixed meals using the white bread glycemic index. Diabetes Care 1985;8:418–28.[Abstract]
  47. Augustin LS, Gallus S, Negri E, La Vecchia C. Glycemic index, glycemic load and risk of gastric cancer. Ann Oncol 2004;15:581–4.[Abstract/Free Full Text]
  48. Larsson SC, Bergkvist L, Wolk A. Glycemic load, glycemic index and carbohydrate intake in relation to risk of stomach cancer: a prospective study. Int J Cancer 2006;118:3167–9.[Medline]
  49. Silvera SA, Jain M, Howe GR, Miller AB, Rohan TE. Glycaemic index, glycaemic load and ovarian cancer risk: a prospective cohort study. Public Health Nutr 2007;10:1076–81.[Medline]
  50. Silvera SA, Rohan TE, Jain M, Terry PD, Howe GR, Miller AB. Glycaemic index, glycaemic load and risk of endometrial cancer: a prospective cohort study. Public Health Nutr 2005;8:912–9.[Medline]
  51. Augustin LS, Gallus S, Bosetti C, et al. Glycemic index and glycemic load in endometrial cancer. Int J Cancer 2003;105:404–7.[Medline]
  52. Larsson SC, Friberg E, Wolk A. Carbohydrate intake, glycemic index and glycemic load in relation to risk of endometrial cancer: a prospective study of Swedish women. Int J Cancer 2007;120:1103–7.[Medline]
  53. Franceschi S, Dal Maso L, Augustin L, et al. Dietary glycemic load and colorectal cancer risk. Ann Oncol 2001;12:173–8.[Abstract/Free Full Text]
  54. Levi F, Pasche C, Lucchini F, Bosetti C, La Vecchia C. Glycaemic index, breast and colorectal cancer. Ann Oncol 2002;13:1688–9.[Free Full Text]
  55. Strayer L, Jacobs DRJ, Schairer C, Schatzkin A, Flood A. Dietary carbohydrate, glycemic index, and glycemic load and the risk of colorectal cancer in the BCDDP cohort. Cancer Causes Control 2007;18:853–63.[Medline]
  56. Weijenberg MP, Mullie PF, Brants HA, Heinen MM, Goldbohm RA, van den Brandt PA. Dietary glycemic load, glycemic index and colorectal cancer risk: results from the Netherlands Cohort Study. Int J Cancer 2008;122:620–9.[Medline]
  57. Augustin LS, Dal Maso L, La Vecchia C, et al. Dietary glycemic index and glycemic load, and breast cancer risk: a case-control study. Ann Oncol 2001;12:1533–8.[Abstract/Free Full Text]
  58. Sieri S, Pala V, Brighenti F, et al. Dietary glycemic index, glycemic load, and the risk of breast cancer in an Italian prospective cohort study. Am J Clin Nutr 2007;86:1160–6.[Abstract/Free Full Text]
  59. Lajous M, Willett W, Lazcano-Ponce E, Sanchez-Zamorano LM, Hernandez-Avila M, Romieu I. Glycemic load, glycemic index, and the risk of breast cancer among Mexican women. Cancer Causes Control 2005;16:1165–9.[Medline]
  60. Silvera SA, Jain M, Howe GR, Miller AB, Rohan TE. Dietary carbohydrates and breast cancer risk: a prospective study of the roles of overall glycemic index and glycemic load. Int J Cancer 2005;114:653–8.[Medline]
  61. Higginbotham S, Zhang ZF, Lee IM, Cook NR, Buring JE, Liu S. Dietary glycemic load and breast cancer risk in the Women's Health Study. Cancer Epidemiol Biomarkers Prev 2004;13:65–70.[Abstract/Free Full Text]
  62. McCann SE, McCann WE, Hong CC, et al. Dietary patterns related to glycemic index and load and risk of premenopausal and postmenopausal breast cancer in the Western New York Exposure and Breast Cancer Study. Am J Clin Nutr 2007;86:465–71.[Abstract/Free Full Text]
  63. Terry PD, Jain M, Miller AB, Howe GR, Rohan TE. Glycemic load, carbohydrate intake, and risk of colorectal cancer in women: a prospective cohort study. J Natl Cancer Inst 2003;95:914–6.[Abstract/Free Full Text]
  64. Higginbotham S, Zhang ZF, Lee IM, et al. Dietary glycemic load and risk of colorectal cancer in the Women's Health Study. J Natl Cancer Inst 2004;96:229–33.[Abstract/Free Full Text]
  65. Larsson SC, Giovannucci E, Wolk A. Dietary carbohydrate, glycemic index, and glycemic load in relation to risk of colorectal cancer in women. Am J Epidemiol 2007;165:256–61.[Abstract/Free Full Text]
  66. Silvera SA, Rohan TE, Jain M, Terry PD, Howe GR, Miller AB. Glycemic index, glycemic load, and pancreatic cancer risk (Canada). Cancer Causes Control 2005;16:431–6.[Medline]
  67. Patel AV, McCullough ML, Pavluck AL, Jacobs EJ, Thun MJ, Calle EE. Glycemic load, glycemic index, and carbohydrate intake in relation to pancreatic cancer risk in a large US cohort. Cancer Causes Control 2007;18:287–94.[Medline]
  68. Augustin LS, Gallus S, Franceschi S, et al. Glycemic index and load and risk of upper aero-digestive tract neoplasms (Italy). Cancer Causes Control 2003;14:657–62.[Medline]
  69. Cho E, Spiegelman D, Hunter DJ, Chen WY, Colditz GA, Willett WC. Premenopausal dietary carbohydrate, glycemic index, glycemic load, and fiber in relation to risk of breast cancer. Cancer Epidemiol Biomarkers Prev 2003;12:1153–8.[Abstract/Free Full Text]
  70. Augustin LS, Galeone C, Dal Maso L, et al. Glycemic index, glycemic load and risk of prostate cancer. Int J Cancer 2004;112:446–50.[Medline]
  71. Randi G, Ferraroni M, Talamini R, et al. Glycemic index, glycemic load and thyroid cancer risk. Ann Oncol 2009;19:380–3 (Epub 2007 Oct 19).
  72. Jonas CR, McCullough ML, Teras LR, Walker-Thurmond KA, Thun MJ, Calle EE. Dietary glycemic index, glycemic load, and risk of incident breast cancer in postmenopausal women. Cancer Epidemiol Biomarkers Prev 2003;12:573–7.[Abstract/Free Full Text]
  73. Augustin LS, Polesel J, Bosetti C, et al. Dietary glycemic index, glycemic load and ovarian cancer risk: a case-control study in Italy. Ann Oncol 2003;14:78–84.[Abstract/Free Full Text]
  74. Flood A, Peters U, Jenkins DJ, et al. Carbohydrate, glycemic index, and glycemic load and colorectal adenomas in the Prostate, Lung, Colorectal, and Ovarian Screening Study. Am J Clin Nutr 2006;84:1184–92.[Abstract/Free Full Text]
  75. Favero A, Salvini S, Russo A, et al. Sources of macro- and micronutrients in Italian women: results from a food frequency questionnaire for cancer studies. Eur J Cancer Prev 1997;6:277–87.[Medline]
  76. Slimani N, Fahey M, Welch AA, et al. Diversity of dietary patterns observed in the European Prospective Investigation into Cancer and Nutrition (EPIC) project. Public Health Nutr 2002;5:1311–28.[Medline]
  77. Wirfalt E, McTaggart A, Pala V, et al. Food sources of carbohydrates in a European cohort of adults. Public Health Nutr 2002;5:1197–215.[Medline]
  78. Giovannucci E. Nutrition, insulin, insulin-like growth factors and cancer. Horm Metab Res 2003;35:694–704.[Medline]
  79. Goodyear LJ, Kahn BB. Exercise, glucose transport, and insulin sensitivity. Annu Rev Med 1998;49:235–61.[Medline]
  80. Key TJ, Spencer EA. Carbohydrates and cancer: an overview of the epidemiological evidence. Eur J Clin Nutr 2007;61(suppl):S112–21.[Medline]
  81. Lin J, Ridker PM, Rifai N, et al. A prospective study of hemoglobin A1c concentrations and risk of breast cancer in women. Cancer Res 2006;66:2869–75.[Abstract/Free Full Text]
  82. Brand-Miller JC, Liu V, Petocz P, Baxter RC. The glycemic index of foods influences postprandial insulin-like growth factor-binding protein responses in lean young subjects. Am J Clin Nutr 2005;82:350–4.[Abstract/Free Full Text]
  83. Perks CM, Newcomb PV, Norman MR, Holly JM. Effect of insulin-like growth factor binding protein-1 on integrin signalling and the induction of apoptosis in human breast cancer cells. J Mol Endocrinol 1999;22:141–50.[Abstract]
  84. Kaaks R, Toniolo P, Akhmedkhanov A, et al. Serum C-peptide, insulin-like growth factor (IGF)-I, IGF-binding proteins, and colorectal cancer risk in women. J Natl Cancer Inst 2000;92:1592–600.[Abstract/Free Full Text]
  85. Mann J. Dietary carbohydrate: relationship to cardiovascular disease and disorders of carbohydrate metabolism. Eur J Clin Nutr 2007;61(suppl):S100–11.[Medline]
  86. Livesey G, Taylor R, Hulshof T, Howlett J. Glycemic response and health a systematic review and meta-analysis: relations between dietary glycemic properties and health outcomes. Am J Clin Nutr 2008;87(suppl):258S–68S.[Abstract/Free Full Text]
  87. Monro JA, Shaw M. Glycemic impact, glycemic glucose equivalents, glycemic index, and glycemic load: definitions, distinctions, and implications. Am J Clin Nutr 2008;87(suppl):237S–43S.[Abstract/Free Full Text]
Received for publication November 27, 2007. Accepted for publication March 5, 2008.




This article has been cited by other articles:


Home page
 Ann OncolHome page
C. Galeone, C. Pelucchi, L. D. Maso, E. Negri, R. Talamini, M. Montella, V. Ramazzotti, R. Bellocco, S. Franceschi, and C. La Vecchia
Glycemic index, glycemic load and renal cell carcinoma risk
Ann. Onc., November 1, 2009; 20(11): 1881 - 1885.
[Abstract] [Full Text] [PDF]

Home page
 Ann OncolHome page
M. Rossi, L. Lipworth, L. D. Maso, R. Talamini, M. Montella, J. Polesel, J. K. McLaughlin, M. Parpinel, S. Franceschi, P. Lagiou, et al.
Dietary glycemic load and hepatocellular carcinoma with or without chronic hepatitis infection
Ann. Onc., October 1, 2009; 20(10): 1736 - 1740.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
A. H Wu, M. C Yu, C.-C. Tseng, F. Z Stanczyk, and M. C Pike
Dietary patterns and breast cancer risk in Asian American women
Am. J. Clinical Nutrition, April 1, 2009; 89(4): 1145 - 1154.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Epidemiol. Biomarkers Prev.Home page
L. Jiao, A. Flood, A. F. Subar, A. R. Hollenbeck, A. Schatzkin, and R. Stolzenberg-Solomon
Glycemic Index, Carbohydrates, Glycemic Load, and the Risk of Pancreatic Cancer in a Prospective Cohort Study
Cancer Epidemiol. Biomarkers Prev., April 1, 2009; 18(4): 1144 - 1151.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Clin. Nutr.Home page
H. G Mulholland, L. J Murray, C. R Cardwell, and M. M Cantwell
Glycemic index, glycemic load, and risk of digestive tract neoplasms: a systematic review and meta-analysis
Am. J. Clinical Nutrition, February 1, 2009; 89(2): 568 - 576.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.
Agricola
Right arrow Articles by Gnagnarella, P.
Right arrow Articles by Maisonneuve, P.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS