Dietary pattern as a predictor of colorectal cancer among general health population in Arusha Tanzania: A population based descriptive study

Proper diet is important in preventing many diseases, and colorectal cancer is no exception. The aim of this study was to identify major dietary patterns among the general population in Arusha Tanzania to determine whether diet is one of the predictors contributing to the observed pattern and distribution of colorectal cancer in Tanzania. A population based cross-sectional study recruited a sample of self-reported healthy individuals residing in four wards of the City of Arusha, Tanzania. A total of 549 participants were recruited on a voluntary basis. The Food Frequency Questionnaire and the World Health Organization (WHO) Step® survey tool were used to collect data. Factor analysis, Pearson correlation (Pearson’s r), and logistic regression were used to analyze the data.Two major dietary patterns, namely “healthy” and “western”, and one minor pattern existed among the study population. The "healthy" pattern was generally associated with females (56.2%, p=0.074), people with primary level of education (62.7%, p=0.667), age category of 25 to 44 (66.3%, p= 0.370), normal range body mass index (BMI) (42.4%, p=0.967), self-employed (78.5%), non-smokers (86.6%) and non-alcohol drinkers (51%), although the differences were not statistically significant. "Western" dietary pattern adherence was associated with area of residence (p=0.0001), gender (p=0.003) and BMI status (p=0.04) in univariate analysis. In multivariate analysis, higher odds were observed in individuals aged 25 to 34 (OR=1.104, 95%, CI(0.537-2.2267) and 45 to 54 OR=1.091, 95%, CI(0.521-2.283), alcohol drinkers (OR=1.2, 95%, CI(0.767-1.877), people with college or high levels of education (OR=0. 853, 95%, CI(0.260-2.803) and OR=0.550, 95%,CI(0. 159-1.897), smokers (OR=1.030, 95%, CI(0.519-2.044) and overweight or obese (OR=2.676, 95%, CI(0.981-7.298) and OR=2.045, 95%, CI(0.767-5.454). These data support our previous hypothesis that diet could be an important potential predictor of the previously observed pattern and distribution of colorectal cancer in Tanzania.