• breast cancer;
  • dietary fibre;
  • carbohydrate;
  • glycaemic index;
  • glycaemic load;
  • post-menopausal women;
  • cohort study;
  • Australia


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

Evidence that the insulin pathway may be involved in breast carcinogenesis has increased the interest in dietary factors that influence insulin secretion and resistance. We investigated dietary carbohydrate, fibre, glycaemic index (GI) and glycaemic load (GL) in a prospective study of 324 breast cancers diagnosed in 12,273 post-menopausal women. Although an increase of 1 standard deviation in carbohydrate was marginally associated with risk of breast cancer, relative risk (RR) 1.31 (95% CI, 0.98, 1.75), there were no significant associations with fibre, 1.08 (0.92, 1.26), GI, 0.98 (0.88, 1.10) or GL, 1.19 (0.93, 1.52) or with carbohydrate foods (bread, rice, pasta). The RR for carbohydrate and localized disease was elevated, 1.40 (1.02, 1.92), but like those for fibre, GI and GL did not differ significantly between localized and non-localized disease. RRs for grade I, but not grade II or III, tumours were elevated for fibre, 1.38 (1.08, 1.75), carbohydrate, 1.56 (1.08, 2.25) and GL, 1.41 (1.01, 1.98) but not for GI, 0.84 (0.65, 1.09). The RRs for fibre and oestrogen receptor (ER) positive (+) and progesterone receptor (PR) positive (+) tumours, 1.36 (1.10, 1.67), differed significantly from those for ER positive (+) and PR negative (−) tumours, 1.01 (0.61, 1.69) and ER−/PR− tumours, 0.65 (0.43, 0.99), p = 0.005. Our data do not support a strong role for GI and GL in breast carcinogenesis but suggest that increased intake of fibre and carbohydrate may be associated with the diagnosis of cancers of more favourable prognosis. © 2005 Wiley-Liss, Inc.

There is mounting evidence that diets characterised by energy imbalance and low physical activity may be important to the risk of many chronic diseases, including breast cancer, and this risk might be mediated by insulin and insulinlike growth factors (IGFs).1, 2, 3 The amount and type of dietary carbohydrate expressed as glycaemic index (GI), or glycaemic load (GL), contribute to variation in blood glucose levels and the body's insulin response.4 Chronic hyperinsulinaemia consequent upon particular dietary GI and GL may lead to increased levels of circulating IGFs that may increase the cancer risk by encouraging cellular proliferation or by other as yet unknown mechanisms.5

Although positive associations between the risk of post-menopausal breast cancer and intakes of GI and GL have been reported by retrospective and case-control studies,6, 7, 8, 9 5 prospective studies have provided little supportive evidence with respect to intakes in adult life.10, 11, 12, 13, 14 We investigated associations between dietary carbohydrates, GI and GL, and dietary fibre, and risk of invasive breast cancer in a prospective study of post-menopausal women with a wide range of relevant dietary exposures and detailed tumour characteristics.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

Study cohort and subjects

In 1990–94, the Melbourne Collaborative Cohort Study (MCCS) recruited 41,528 people (24,479 women) from the Melbourne metropolitan area. Most participants (99.3%) were aged 40–69 years. Migrants from Italy and Greece were deliberately recruited to increase the range of dietary exposures available.15 It was approved by The Cancer Council Victoria's Human Research Ethics Committee.

Subjects were excluded from analysis if they had a confirmed invasive breast cancer prior to baseline (386 women) and were pre-menopausal or of unknown menopausal status at baseline (10,374). A further 1,446 post-menopausal women were excluded because they had a self-reported heart attack, diabetes and angina at baseline; had missing data for the food frequency questionnaire (FFQ); or extreme values of total energy intake (<1st percentile and >99th percentile). The analysis was based on a total of 12,273 post-menopausal women.

Dietary assessment

Dietary information was collected at baseline using a 121 item, self-administered FFQ, specifically developed for the MCCS.16 Glycaemic index (GI) is a method of ranking foods on the basis of the blood glucose response to a given amount of carbohydrate from that food. GI values of individual food items, relative to glucose, were obtained from the 2002 International table of GI and glycaemic load (GL) values.17 Where there was more than one value, GI values were averaged, with preference being given to Australian figures. Dietary GL was computed by summing the product of carbohydrate intake from each food with the GI for that food. GL was divided by total carbohydrate intake to obtain dietary GI, that is, an average of individual food GI values, weighted according to their contribution to carbohydrate intake. Further detail on the GI calculations has been published elsewhere.18 Alcoholic beverages were not included in the overall GI. Intakes of energy, fibre (total and separately from cereals, fruits, vegetables, legumes and potatoes), total carbohydrate, total starch and total sugars were computed using Australian food composition tables.19 The food groups investigated were total cereal products, breakfast cereal, bread, wholemeal bread, white bread, rice, and pasta and noodles as number of times eaten per week.

Other measures

At interview, questions were asked about reproductive history, country of birth, alcohol intake, physical activity, highest level of education and use of hormone replacement therapy (HRT) and oral contraceptives (OC).

Ascertainment of invasive breast cancer cases

All women consented to us accessing their medical records, and their cancers were ascertained by the Victorian Cancer Registry, which has complete coverage of the cohort. Altogether, 324 invasive breast cancers were identified and their tumour, nodes and metastases stage was grouped as localized (stage I) and non-localized (stages II–IV). Tumour grade and oestrogen receptor (ER) and progesterone receptor (PR) status were obtained from pathology reports.

Statistical analysis

Follow-up began at baseline and continued until a diagnosis of breast cancer, death, emigration from Victoria or 30 June 2002, whichever came first. By 30 June 2002, 161 (1.3%) post-menopausal women had left Victoria. Cox's proportional hazards regression, with age as the time axis, was used to derive relative risk (RR) for each nutrient and food group measured at baseline. Nutrients were treated as continuous variables, the RR being based on a one standard deviation increase. Food group variables were treated as pseudo-continuous variables by assigning the median value within each quartile grouping. The RRs were based on an increase of 1 serve/week.

Country of birth and total energy intake (kJ/day) were included in all models. Potential confounders were included in final analyses if they changed the RR of any nutrient by at least 5%. These included parity, age at first full-term pregnancy (defined as a live birth or pregnancy with a gestation >24 weeks), cumulative lactation (months), HRT, OC use, age at menarche and menopause, maternal family history of cancer, physical activity, alcohol intake (g/day), highest level of education, height, body weight, body mass index, waist–hip ratio, multi-vitamin supplement use and total fat intake (g/day). The final model included country of birth, total energy intake and HRT.

Heterogeneity in the RRs (estimated by the odds ratio) for stage (localized, non-localized), grade (I–III) and ER/PR status (ER+, ER−, PR+, PR− considered separately and jointly) was investigated using polytomous logistic regression adjusted for age (continuous), country of birth, total energy intake and HRT.

All analyses were computed using STATA statistical software (Version 7.0. Stata Corporation, College Station, TX, USA).


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

Altogether, 324 invasive breast cancers were identified during an average of 9.1 years' follow-up. The stage was available for 299 (92.3%) cancers, 172 (58%) localized and 127 (42%) non-localized. Grade was known for 282 (87%) cancers, 64 (23%), 126 (45%) and 92 (33%) being grade I, II and III respectively. ER/PR status was available for 258 (79.6%) cancers, of which 154 (60%) were ER+/PR+, 39 (15%) ER+/PR−, 13 (5%) ER−/PR+ and 52 (20%) ER−/PR−.

The median and inter-quintile cut points were 28.9 (21.1, 38.5) g/day for total fibre intake, 219 (157, 302) g/day for carbohydrate, 48.8 (45.6, 52.7) for GI and 108.2 (76.9, 149.1) for GL (Table I). The RRs and 95% confidence limits for one standard deviation increase in intake were as follows: total fibre 1.08 (0.92, 1.26) p = 0.36, total carbohydrate 1.31 (0.98, 1.75) p = 0.07, GI 0.98 (0.88, 1.10) p = 0.74 and GL 1.19 (0.93, 1.52) p = 0.17. None of the RRs for different sources of dietary fibre or for the 7 food groups differed significantly from unity (Table II).

Table I. Demographic Characteristics and Nutrient Intake for the 12,273 Post-Menopausal Women
Demographic characteristic and nutrient intake 
Country of birth (%)
Hormone replacement therapy (%)
Education (%)
 Primary school24.2
 Some high school45.9
 Completed high school16.7
Alcohol intake
 Low (<20 g/day)90.0
 Medium (20–39 g/day)7.8
 High (40+ g/day)2.2
Mean BMI ± SD (kg/m2)27.1 ± 4.8
Median parity [inter-quintile range]2 [1, 4]
Age at menarche (<12 years) (%)34.1
Age at first term pregnancy (years) (%) 
 No term pregnancy12.1
Nutrient intake 
 Median energy intake (kJ/d) [inter-quintile range]8100 [6118, 10634]
 Median fibre intake (g/d) [inter-quintile range]28.9 [21.1, 38.5]
 Median carbohydrate intake (g/d) [inter-quintile range]219.0 [156.9, 301.5]
 Median starch intake (g/d) [inter-quintile range]104.9 [74.5, 144.1]
 Median sugar intake (g/d) [inter-quintile range]108.0 [71.9, 162.8]
 Median glycaemic index [inter-quintile range]48.8 [45.6, 52.7]
 Median glycaemic load [inter-quintile range]108.2 [76.9, 149.1]
Table II. Adjusted Relative Risks1 for the Association Between Glycaemic Index, Glycaemic Load, Nutrient and Food Group Intake, and Breast Cancer Risk in Localized and Non-Localized Tumours
 All cases (n = 324)Non-localized (n = 127)Localized (n = 172)p-value2
  • 1

    Adjusted for age at attendance, country of birth, total energy intake and hormone replacement therapy. Relative risks represent the risk associated with an increase of 1 SD for glycaemic index, glycaemic load and nutrients; and the risk associated with an increase of 1 unit in the number of times/week for the food groups.

  • 2

    From likelihood ratio test investigating heterogeneity in the relative risks for stage of breast cancer.

  • 3

    Values in parentheses indicate 95% confidence interval.

  • 4

    Treated as a pseudo-continuous variable.

Nutrient intake (g/day)
Total fibre1.08 (0.92, 1.26)31.10 (0.90, 1.36)1.10 (0.91, 1.33)0.97
 From cereals1.08 (0.95, 1.23)1.04 (0.85, 1.26)1.16 (0.99, 1.36)0.33
 From fruit1.00 (0.88, 1.13)1.12 (0.95, 1.32)0.91 (0.76, 1.09)0.08
 From vegetables1.07 (0.95, 1.20)1.01 (0.82, 1.23)1.11 (0.96, 1.29)0.40
 From legumes0.97 (0.86, 1.11)0.99 (0.82, 1.20)0.97 (0.81, 1.16)0.84
 From potatoes0.97 (0.85, 1.11)1.00 (0.82, 1.22)0.95 (0.79, 1.15)0.71
Total carbohydrate1.31 (0.98, 1.75)1.18 (0.84, 1.66)1.40 (1.02, 1.92)0.16
Total starch1.14 (0.95, 1.38)1.06 (0.83, 1.36)1.24 (1.00, 1.54)0.18
Total sugars1.07 (0.89, 1.28)0.96 (0.75, 1.23)1.12 (0.90, 1.38)0.21
Food group intake (times/week)4
All cereal products1.01 (1.00, 1.02)1.00 (0.99, 1.02)1.01 (1.00, 1.03)0.33
 Breakfast cereal1.01 (0.97, 1.05)0.99 (0.93, 1.05)1.02 (0.97, 1.07)0.53
 Total bread1.01 (0.99, 1.03)1.00 (0.98, 1.03)1.02 (1.00, 1.04)0.34
  Wholemeal bread1.00 (0.98, 1.02)1.00 (0.97, 1.02)1.01 (0.99, 1.03)0.44
  White bread1.00 (0.99, 1.02)1.00 (0.98, 1.03)1.01 (0.99, 1.03)0.84
 Rice0.96 (0.88, 1.05)0.98 (0.86, 1.13)0.93 (0.83, 1.05)0.56
 Pasta and noodles1.03 (0.92, 1.15)1.07 (0.90, 1.26)0.95 (0.82, 1.11)0.29
Glycaemic index and load
Glycaemic index0.98 (0.88, 1.10)0.99 (0.82, 1.18)0.96 (0.83, 1.12)0.85
Glycaemic load1.19 (0.93, 1.52)1.07 (0.80, 1.44)1.25 (0.95, 1.64)0.21

Although there was no evidence of heterogeneity in the RRs for localized and non-localized breast cancer (Table II), the RR for carbohydrate intake and localized disease was elevated, 1.40 (1.02, 1.92).

Similarly, although no evidence of heterogeneity by tumour grade was observed for any nutrient or food group, the RRs for fibre, carbohydrate and GL, but not GI, were all significantly elevated for low-grade tumours (Table III).

Table III. Adjusted Relative Risks1 for the Association Between Glycaemic Index, Glycaemic Load, Nutrient and Food Group Intake, and Breast Cancer Risk by Tumour Grade
 Grade I (n = 64)Grade II (n = 126)Grade III (n = 92)p-value2
  • 1

    Adjusted for age at attendance, country of birth, total energy intake and hormone replacement therapy. Relative risks represent the risk associated with an increase of 1 SD for glycaemic index, glycaemic load and nutrients; and the risk associated with an increase of 1 unit in the number of times/week for the food groups.

  • 2

    From likelihood ratio test investigating heterogeneity in the relative risks for tumour grade of breast cancer.

  • 3

    Values in parentheses indicate 95% confidence interval.

  • 4

    Treated as a pseudo-continuous variable.

Nutrient intake (g/day)
Total fibre1.38 (1.08,1.75)31.08 (0.87,1.34)1.03 (0.81,1.32)0.14
 From cereals1.28 (1.04,1.58)1.15 (0.96,1.37)1.01 (0.81,1.27)0.28
 From fruit1.16 (0.94,1.44)0.93 (0.76,1.14)0.96 (0.76,1.21)0.31
 From vegetables1.23 (1.03,1.47)1.06 (0.88,1.28)1.03 (0.82,1.29)0.42
 From legumes1.09 (0.94,1.26)0.93 (0.74,1.17)0.97 (0.76,1.23)0.52
 From potatoes0.86 (0.61,1.21)1.01 (0.83,1.22)1.03 (0.83,1.27)0.63
Total carbohydrate1.56 (1.08,2.25)1.31 (0.93,1.84)1.28 (0.89,1.84)0.41
Total starch1.30 (0.98,1.74)1.26 (1.00,1.60)1.14 (0.87,1.49)0.65
Total sugars1.22 (0.93,1.60)0.96 (0.75,1.23)1.01 (0.77,1.32)0.27
Food group intake (times/week)4
Total cereal products1.02 (1.00,1.04)1.01 (0.99,1.02)1.00 (0.99,1.02)0.32
 Breakfast cereal1.05 (0.96,1.14)0.99 (0.93,1.05)0.99 (0.93,1.07)0.54
 Total bread1.02 (0.98,1.05)1.03 (1.00,1.05)0.99 (0.96,1.02)0.07
  Wholemeal bread1.02 (0.98,1.06)1.02 (0.99,1.04)0.98 (0.95,1.01)0.13
  White bread0.99 (0.95,1.02)1.01 (0.99,1.03)1.01 (0.99,1.04)0.41
 Rice0.98 (0.81,1.19)0.99 (0.86,1.13)0.98 (0.84,1.15)0.99
 Pasta and noodles1.01 (0.80,1.28)0.97 (0.82,1.16)1.12 (0.93,1.36)0.52
Glycaemic index and load
Glycaemic index0.84 (0.65,1.09)1.11 (0.93,1.32)0.93 (0.76,1.15)0.18
Glycaemic load1.41 (1.01,1.98)1.23 (0.91,1.66)1.16 (0.84,1.61)0.49

Weak heterogeneity in the RRs for the ER/PR subtypes was observed for total fibre intake (Table IV). An increase in fibre intake was associated with an increased risk of ER+/PR+ tumours, null risk of ER+/PR− tumours and a reduced risk of ER−/PR− tumours, p = 0.005. When analysed individually as ER+ or ER−, the RRs for fibre were 1.18 (0.98, 1.41) and 0.89 (0.66, 1.21) respectively (p = 0.07). Similarly the RRs for fibre and PR+ and PR− tumours were 1.23 (1.02, 1.48) and 0.88 (0.67, 1.14) respectively, p = 0.01.

Table IV. Adjusted Relative Risks1 for the Association Between Glycaemic Index, Glycaemic Load, Nutrient and Food Group Intake, and Breast Cancer Risk by ER/PR Status
 ER+/PR+ (n = 154)ER+/PR− (n = 39)ER−/PR− (n = 52)p-value2
  • 1

    Adjusted for age at attendance, country of birth, total energy intake and hormone replacement therapy. Relative risks represent the risk associated with an increase of 1 SD for glycaemic index, glycaemic load and nutrients; and the risk associated with an increase of 1 unit in the number of times/week for the food groups. ER−/PR+ (N = 13) were excluded because of small numbers.

  • 2

    From likelihood ratio test investigating heterogeneity in the relative risks for ER/PR status of breast cancer.

  • 3

    Values in parentheses indicate 95% confidence interval.

  • 4

    Treated as a pseudo-continuous variable.

Nutrient intake (g/day)
Total fibre1.36 (1.10,1.67)31.01 (0.61,1.69)0.65 (0.43,0.99)0.005
 From cereals1.17 (0.98,1.39)1.24 (0.83,1.86)0.78 (0.55,1.11)0.09
 From fruit1.15 (1.00,1.34)0.82 (0.53,1.28)0.78 (0.55,1.11)0.05
 From vegetables1.13 (0.97,1.32)1.14 (0.81,1.61)0.83 (0.58,1.19)0.24
 From legumes1.01 (0.86,1.18)0.95 (0.61,1.46)0.87 (0.60,1.27)0.74
 From potatoes0.98 (0.82,1.18)0.79 (0.48,1.32)1.07 (0.85,1.36)0.54
Total carbohydrate1.38 (0.91,2.09)1.91 (0.75,4.86)0.88 (0.45,1.71)0.36
Total starch1.07 (0.82,1.41)1.60 (0.91,2.84)0.80 (0.50,1.28)0.18
Total sugars1.13 (0.87,1.46)1.07 (0.58,1.95)1.06 (0.69,1.63)0.96
Food group intake (times/week)4
All cereal products1.01 (1.00,1.02)1.01 (0.98,1.04)1.00 (0.98,1.03)0.88
 Breakfast cereal1.03 (0.97,1.09)0.97 (0.87,1.08)0.97 (0.88,1.06)0.41
 Total bread1.01 (0.99,1.04)1.01 (0.97,1.06)0.98 (0.94,1.02)0.28
  Wholemeal bread1.01 (0.99,1.04)1.02 (0.97,1.07)0.98 (0.94,1.03)0.45
  White bread1.00 (0.98,1.02)0.99 (0.94,1.03)1.00 (0.96,1.03)0.90
Rice0.95 (0.83,1.08)0.94 (0.72,1.21)1.04 (0.85,1.27)0.71
Pasta and noodles1.14 (0.98,1.32)0.71 (0.48,1.03)1.03 (0.80,1.34)0.04
Glycaemic index and load
Glycaemic index0.91 (0.77,1.07)0.80 (0.57,1.12)0.98 (0.74,1.29)0.65
Glycaemic load1.11 (0.78,1.59)1.32 (0.60,2.90)0.81 (0.46,1.44)0.55


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References

We found little evidence to support more than subtle, if any, influences of dietary carbohydrate, fibre and GI or GL intakes on post-menopausal invasive breast cancer risk. Our findings are in general agreement with other recent prospective cohort studies that have examined adult exposures and risk of post-menopausal breast cancer.10, 11, 12, 13, 14 On the other hand, the retrospective cohort analysis of the Nurses Health Study did find associations when using estimates of dietary intakes recalled from adolescence.8

The strengths of our prospective study include its reasonable size and length of follow-up, its capture of tumour stage, grade and ER/PR status, and its wide range of dietary exposures due to the inclusion of southern European subjects. The Nurse's Health study, in comparison, is reported to be a very homogeneous group in terms of dietary intake and, therefore, possibly not an ideal one in which to detect modest associations.20 For example, investigators of the EPIC study, comparing their findings of a protective effect of fibre intake on colorectal cancer incidence with the finding of no effect reported by the Nurses' Health Study, cited two main reasons for the difference.21 First was the lower consumption of fibre in the Nurses study wherein intake ranged from 9.8 to 24.9 g/day compared with 12.6–33.1 g/day for EPIC study. Fibre intake from cereals was even lower for the Nurses Study: 4.8 g/day in the highest intake group compared with 4.7 in the lowest intake group from the EPIC study. Second was that participants in the EPIC study also consumed fibre from a greater variety of sources than did participants in the US Nurses Study.21

Our study's weaknesses are common to many other prospective studies and include the possibility that the diet may have changed during the follow-up, leading to increased misclassification of long-term exposures, the large measurement errors associated with estimating food and nutrient intakes, and the under-reporting of food and energy intake, particularly by those at the upper extreme of body mass.22 One possible effect of these is to attenuate risk estimates. The risks associated with dietary factors are likely to be relatively small, and we cannot exclude the possibility that associations with breast cancer may have been found had we been able to measure dietary risk factors more accurately. For example, in a recent study of dietary fat and breast cancer, saturated fat measured with a food diary was associated with breast cancer, while saturated fat measured with a FFQ was not.23 Although we performed sub analyses by stage and ER/PR status because we had collected this information, we recognize that we had limited statistical power with the small number of cases (n = 324) currently available.

Because the FFQ was developed before GI values for a wide range of foods were available, it was not designed specifically for GI calculations, and some details are lacking in the description of carbohydrate foods, making it difficult to accurately attribute the GI values. The GI and GL values calculated from the FFQ have not been evaluated, but we have previously shown that the GI is associated with incidence of type 2 diabetes in the cohort, using the same data.18

The evidence from prospective cohort studies of any effect of these dietary factors is sparse, with only limited evidence for fibre and GI. In regard to GI, an analysis of 1,461 breast cancers (including in situ tumours) during 16.6 years of follow-up of 49,613 women in The Canadian National Breast Screening Study comparing highest and lowest quintiles of GI found an RR for post-menopausal breast cancer of 1.87 (1.18, 2.97), p trend = 0.01,14 the association being slightly stronger for women who reported no vigorous activity, ever use of HRT and normal weight.14 The Danish “Diet, Cancer and Health” study examined carbohydrate, GI and GL intakes of 634 incident breast cancers (with ER/PR status) diagnosed among 23,870 women and found very little, the only exception being a borderline positive association between GI and ER− cancers, RR 1.46 (1.01, 2.11) per 10 units/day.13 The most recent analysis of the Nurses Health Study found no association between dietary GI or GL and risk of either pre- or post-menopausal breast cancer;12 the analysis of 1,442 breast cancers (including 174 in situ) in the CPSII cohort of 63,307 women followed for 5 years10 found no association with either GI or GL and no modification by stage or body size; and the analysis of 946 breast cancers diagnosed during 6.8 years' follow-up of 39,876 participants of the Women's Health Study found no associations between GI or GL and post-menopausal breast cancer.11 The Danish report is consistent with being a chance finding, and the Canadian study that is based on a screening program may have been subject to selection and detection biases.

The strongest association between risk of breast cancer and fibre is from the Malmo Diet and Cancer Cohort. The analysis of 342 cases of breast cancer (including 30 in situ cases) from the follow-up of 11,726 post-menopausal women found evidence of a protective effect of fibre, RR 0.58 (0.40, 0.84) comparing highest and lowest quintiles of intake.24 The only other study to suggest any possible protective effect of fibre was an analysis of 4,092 invasive breast cancers diagnosed in 88,678 participants of the Nurses Health Study, which found an RR of 0.68 (0.43, 1.06) for a small minority of women (<1%) consuming >30 g/day, compared with <10 g/day.12 An earlier analysis of this cohort had shown no effect of fibre intake,25 and the recent report is consistent with a chance observation. The Netherlands Cohort Study of 650 incident breast cancers diagnosed during 4.3 years' follow-up of 62,573 women also reported no association with fibre intake,26 and the Canadian National Breast Screening Study of 2,536 breast cancers (including in situ tumours) during 16.2 years of follow-up of 89,835 women also found no association with total fibre or its sub-fractions.27

The reports of a protective effect of fibre are in contrast to our study that showed it to increase the risk of low grade or ER+/PR+ cancer. In our study, fibre intake was highly correlated with carbohydrate and GL intakes, making it difficult to separate their putative effects. We considered that the effect of fibre observed in the Malmo cohort24 might have been due to a larger range or level of intake, but the medians and inter-quintile ranges for Malmo and the Nurses Health Study12 were very similar, 18.3 (12.5–25.9) and 17.4 (12.1–24.8) respectively, and ours were higher 28.9 (21.1, 38.5). Perhaps the precision of the 7 day food histories used in the Malmo study is superior to the 131 item FFQ used in the Nurses Study and the 121 item FFQ used in our study to estimate fibre intake.

Our findings suggest a modest association between fibre, carbohydrate and the incidence of localized, low grade and ER+/PR+ invasive tumours. Such tumours have a very favourable prognosis, and we consider that our findings are possibly due to increased use of mammographic screening by women who also adhere to a high-fibre, high-carbohydrate dietary pattern, but we do not have any information on the women's screening history with which to test this hypothesis. Our findings do not support either fibre, carbohydrate, GI or GL being associated with the risk of aggressive post-menopausal breast cancer.


  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. References
  • 1
    Cohen LA. Dietary fiber and breast cancer. Anticancer Res 1999; 19: 36858.
  • 2
    Stoll BA. Western nutrition and the insulin resistance syndrome: a link to breast cancer. Eur J Clin Nutr 1999; 53: 837.
  • 3
    Brand-Miller JC. Glycemic load and chronic disease. Nutr Rev 2003; 61: S49S55.
  • 4
    Jenkins DJ, Kendall CW, Augustin LS, Franceschi S, Hamidi M, Marchie A, Jenkins AL, Axelsen M. Glycemic index: overview of implications in health and disease. Am J Clin Nutr 2002; 76: 266SS73.
  • 5
    Renehan AG, Zwahlen M, Minder C, O'Dwyer ST, Shalet SM, Egger M. Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: systematic review and meta-regression analysis. Lancet 2004; 363: 134653.
  • 6
    Augustin LS, Dal Maso L, La Vecchia C, Parpinel M, Negri E, Vaccarella S, Kendall CW, Jenkins DJ, Francesch S. Dietary glycemic index and glycemic load, and breast cancer risk: a case-control study. Ann Oncol 2001; 12: 15338.
  • 7
    Levi F, Pasche C, Lucchini F, Bosetti C, La Vecchia C. Glycaemic index, breast and colorectal cancer. Ann Oncol 2002; 13: 16889.
  • 8
    Frazier AL, Li L, Cho E, Willett WC, Colditz GA. Adolescent diet and risk of breast cancer. Cancer Causes Control 2004; 15: 7382.
  • 9
    Romieu I, Lazcano-Ponce E, Sanchez-Zamorano LM, Willett W, Hernandez-Avila M. Carbohydrates and the risk of breast cancer among Mexican women. Cancer Epidemiol Biomarkers Prev 2004; 13: 12839.
  • 10
    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: 5737.
  • 11
    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: 6570.
  • 12
    Holmes MD, Liu S, Hankinson SE, Colditz GA, Hunter DJ, Willett WC. Dietary carbohydrates, fiber, and breast cancer risk. Am J Epidemiol 2004; 159: 7329.
  • 13
    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: 1248.
  • 14
    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: 6538.
  • 15
    Giles GG. The Melbourne Study of Diet and Cancer. Proc Nutr Soc Aust 1990; 15: 618.
  • 16
    Ireland P, Jolley D, Giles GG, O'Dea K, Powles J, Rutishauser I, Wahlqvist M, Williams J. Development of the Melbourne FFQ: a food frequency questionnaire for use in an Australian prospective study involving an ethnically diverse cohort. Asia Pac J Clin Nutr 1994; 3: 1931.
  • 17
    Foster-Powell K, Holt SH, Brand-Miller JC. International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 2002; 76: 556.
  • 18
    Hodge AM, English DR, O'Dea K, Giles GG. Glycemic index and dietary fiber and the risk of type 2 diabetes. Diabetes Care 2004; 27: 27016.
  • 19
    Lewis J, Milligan G, Hunt A. NUTTAB95 nutrient data table for use in Australia. Canberra: Australian Government Publishing Service, 1995.
  • 20
    Holmes MD, Hunter DJ, Colditz GA, Stampfer MJ, Hankinson SE, Speizer FE, Rosner B, Willett WC. Association of dietary intake of fat and fatty acids with risk of breast cancer. JAMA 1999; 281: 91420.
  • 21
    Bingham SA, Day NE, Luben R, Ferrari P, Slimani N, Norat T, Clavel-Chapelon F, Kesse E, Nieters A, Boeing H, Tjonneland A, Overvad K, et al. Dietary fibre in food and protection against colorectal cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC): an observational study. Lancet 2003; 361: 1496501.
  • 22
    Subar AF, Kipnis V, Troiano RP, Midthune D, Schoeller DA, Bingham S, Sharbaugh CO, Trabulsi J, Runswick S, Ballard-Barbash R, Sunshine J, Schatzkin A. Using intake biomarkers to evaluate the extent of dietary misreporting in a large sample of adults: the OPEN study. Am J Epidemiol 2003; 158: 113.
  • 23
    Bingham SA, Luben R, Welch A, Wareham N, Khaw K-T, Day N. Are imprecise methods obscuring a relation between fat and breast cancer? Lancet 2003; 362: 2124.
  • 24
    Mattisson I, Wirfalt E, Johansson U, Gullberg B, Olsson H, Berglund G. Intakes of plant foods, fiber and fat and risk of breast cancer—a prospective study in the Malmo Diet and Cancer cohort. Br J Cancer 2004; 90: 1227.
  • 25
    Willett WC, Hunter DJ, Stampfer MJ, Colditz G, Manson JE, Spiegelman D, Rosner B, Hennekens CH, Speizer FE. Dietary fat and fiber in relation to risk of breast cancer. An 8-year follow-up. JAMA 1992; 268: 203744.
  • 26
    Verhoeven DT, Assen N, Goldbohm RA, Dorant E, van 't Veer P, Sturmans F, Hermus RJ, van den Brandt PA. Vitamins C and E, retinol, beta-carotene and dietary fibre in relation to breast cancer risk: a prospective cohort study. Br J Cancer 1997; 75: 14955.
  • 27
    Terry P, Jain M, Miller AB, Howe GR, Rohan TE. No association among total dietary fiber, fiber fractions, and risk of breast cancer. Cancer Epidemiol Biomarkers Prev 2002; 11: 15078.