Figure 2
Linear regression analysis of association between wheat consumption and cancer incidence rates. To determine whether the relationship between wheat consumption and cancer incidence was related to geographic origin of wheat, 2007 wheat consumption data was collected from FAOSTAT, defined as kg wheat products consumed per capita per year, and from the GLOBOCAN resource from 2008, defined as age-standardized rates (ASR) per 100,000 of cancer incidence at all sites excluding non-melanoma skin cancer. ASRs were log10-transformed to satisfy statistical criteria. Countries without data for both parameters were excluded from analyses. (A) All-site cancer incidence ASRs for 165 countries were regressed against yearly wheat consumption, which showed a slightly positive correlation between wheat consumption and cancer incidence. Slope = 0.008940 ± 0.002527; Y-intercept = 2.120 ± 0.02197; R2 = 0.07131; line equation: Log10-transformed ASR = 0.008940*wheat consumption + 2.120. (B) A subset of the global analysis, comprising ASRs for n=19 Near Eastern countries, which are geographically proximate to the origin of wheat, were regressed against wheat consumption data, which showed a slightly negative correlation between wheat consumption and cancer incidence. Slope = −0.006526 ± 0.008315; Y-intercept= 2.187 ± 0.1225; R2 = 0.03497; line equation: Log10-transformed ASR = −0.006526*wheat consumption + 2.187. Analyses were performed using the linear regression analysis function in GraphPad Prism vs. 5.02 (GraphPad Software, San Diego, CA, USA).