Erratum


It has come to our attention that there was an error in the data set that was used for the analyses presented in the paper entitled “ Dietary carbohydrates and breast cancer risk: A prospective study of the roles of overall glycemic index and glycemic load.” The correct number of breast cancer cases included in the analyses should be 2,518, not 1,461.

Given the additional cases, the multivariate analyses examining glycemic load, overall glycemic index, and total intakes of carbohydrates and sugar (separately), did not change materially. However, the results of our stratified analyses changed somewhat. In particular, when risk of breast cancer risk associated with overall glycemic index was examined across strata of menopausal status, there was some suggestion of effect modification. Among premenopausal women, the highest versus lowest quintile level of overall glycemic index was associated with a 21% reduction in risk (HR = 0.79, 95% CI = 0.54 – 1.16, Ptrend = 0.02), while among postmenopausal women the highest versus lowest quintile level of overall glycemic index was associated with a 50% increased risk of breast cancer (HR = 1.50, 95% CI = 0.96– 2.33, Ptrend = 0.06). On formal testing, the interaction between overall glycemic index and menopausal status was statistically significant (χ2(4)= 18.39, P = 0.001). Upon re-examining the associations between glycemic load and overall glycemic index among premenopausal women, stratified by body mass index (BMI) and participation in vigorous physical activity, we found that the results did not change materially given the additional cases. However, among postmenopausal women, we initially reported a statistically significant positive trend between overall glycemic index and breast cancer risk among lean women. Using the corrected data set, we observed no association between overall glycemic index and risk among lean women (Ptrend = 0.99), but found a 2.5-fold increase in risk of breast cancer associated with the highest versus lowest quartile level of overall glycemic index among overweight/obese postmenopausal women (95% CI = 1.27 – 5.01, Ptrend = 0.01). The corrected Table V is below:

Table V. Adjusted* Hazard Ratios (95% Confidence Intervals (CI)) for the Association Between Glycemic Load (Quartiles) and Overall Glycemic Index and Risk of Breast Cancer Among Postmenopausal Women by Levels of Body Mass Index, Vigorous Physical Activity, and Hormone Replacement Therapy Use
No. Cases/ Person yearsBMI (kg/m2)Vigorous Physical ActivityHormone Replacement Therapy Use
< 25≥25p for interactionNoneSomep for interactionNeverEverp for interaction
551 / 167,628441 / 122,229 234 / 80,274424 / 162,767 520 / 153,045480 / 140,324 
  • *

    Multivariable models included age (age at baseline), BMI in kg/m2 (<25, 25–29, ≥30), alcohol (zero plus for levels of intake), use of hormone replacement therapy (never plus four levels of duration), use of oral contraceptives (never plus four levels of duration), parity (zero plus 3 levels), age at menarche (≤12 vs. >12 years of age), age at first live birth (< 25, 25–29, ≥30), first degree relative with breast cancer (yes/no), history of benign breast disease (yes/no), intake of energy (as a continuous variable), energy-adjusted total fiber intake (as a continuous variable), study center, and randomization group.

Glycemic Load (g/day)
<1251.0 (referent)1.0 (referent) 1.0 (referent)1.0 (referent) 1.0 (referent)1.0 (referent) 
125 – 1471.19 (0.85 – 1.68)1.08 (0.74 – 1.58) 0.91 (0.58 – 1.42)1.23 (0.90 – 1.68) 1.10 (0.79 – 1.53)1.13 (0.76 – 1.68) 
148 – 1691.05 (0.73 – 1.50)1.15 (0.79 – 1.68) 1.16 (0.76 – 1.78)1.05 (0.76 – 1.46) 0.95 (0.67 – 1.35)1.31 (0.89 – 1.93) 
>1690.95 (0.66 – 1.38)1.20 (0.81 – 1.77) 0.86 (0.54 – 1.36)1.17 (0.84 – 1.63) 0.97 (0.68 – 1.38)1.15 (0.76 – 1.73) 
 p = 0.55p = 0.330.63p = 0.78p = 0.640.17p = 0.66p = 0.410.58
Overall Glycemic Index
<631.0 (referent)1.0 (referent) 1.0 (referent)1.0 (referent) 1.0 (referent)1.0 (referent) 
64 – 761.11 (0.74 – 1.65)1.56 (0.96 – 2.53) 1.62 (0.92 – 2.84)1.13 (0.78 – 1.63) 1.24 (0.83 – 1.86)1.27 (0.79 – 2.05) 
77 – 921.02 (0.63 – 1.65)2.10 (1.19 – 3.69) 1.88 (0.97 – 3.64)1.17 (0.75 – 1.82) 1.22 (0.75 – 1.99)1.58 (0.91 – 2.75) 
>921.07 (0.58 – 1.96)2.52 (1.27 – 5.01) 1.93 (0.86 – 4.33)1.39 (0.80 – 2.40) 1.12 (0.61 – 2.07)2.20 (1.12 – 4.43) 
 p = 0.99p = 0.010.97p = 0.17p = 0.260.72p = 0.93p = 0.020.56

Ancillary