Our study had some limitations which must
Our study had some limitations which must be taken into account when interpreting the findings. Firstly, the food frequency questionnaire was semi-quantitative, and we did not analyse the data according to the nutrient or absolute food quantity. In addition, familial colorectal cancer syndromes were not excluded, because they SC 236 are not adequately investigated in routine clinical practice in our setting. However, these generally constitute a small proportion of all colorectal cancer cases (5%), and the effect on the results is likely to be minimal . We further minimised the effect of familial colorectal cancer syndromes by adjusting for family history of cancer during analysis. Moreover, environmental factors such as diet influence the development of colorectal cancer even in familial cancer syndromes . Case-control studies of this nature are also susceptible to differential recall between cases and controls. We tried to minimise this by recruiting cases as soon as possible after diagnosis, and at most within 6 months. We did not collect data on physical activity and body mass index (BMI), as we felt these parameters are particularly prone to differential recall in our setting. Unlike controls, cases would have been asked for a subjective estimate of their pre-morbid weight and physical activity. It is possible that this was the confounder for the association between tertiary education and colorectal cancer in this study. It is reasonable to assume those individuals with a tertiary education were more sedentary and had a higher BMI, which increased their colorectal cancer risk. It may be argued that the association with education suggests differential access to healthcare, but there was no evidence for this during recruitment. Finally, although we achieved the desired sample size, the study can still be regarded as small in comparison to similar studies, and this may have masked the potentially deleterious impact of the urban and processed foods patterns.
In conclusion, our findings re-affirms the protective properties of the traditional African diets, and demonstrate that urbanisation and rising incomes are associated with a shift from these protective diets. These findings provide a basis for designing primary intervention strategies for populations undergoing dietary transitions, and adds to the general evidence base on the role of diet in colorectal cancer. The promotion of traditional diets in these populations may slow the rise in colorectal cancer. Moreover, such interventions may have a salutary effect on other non-communicable diseases in particular obesity and diabetes mellitus, which to an extent, have similar risk factors to colorectal cancer [25,26]. This may be vital in sub-Saharan Africa, where strategies based on screening are unlikely to be feasible or cost-effective.
Authorship contribution statement
Declarations of interest
Acknowledgement This work was supported by the Wellcome Trust through the Southern African Consortium for Research Excellence (SACORE) initiative.
Introduction Evidence is conflicting regarding the association between glycemic-control (glucose or HbA1c blood-levels) and cancer-risk. A meta-analysis of 4 large randomized controlled trials did not find intensive glycemic-control to be associated with lower cancer-risk . A meta-analysis including 19 studies that compared persons with high versus low levels of serum glucose (cut-off > 6.1 mmol/L) showed serum glucose to have a pooled relative risk (RR) of 1.32 (95%CI:1.20–1.45) for cancer-incidence, . Several studies have investigated associations of hyperglycemia with site-specific cancers. A Japanese case-control study found no association between blood-glucose and the risk of gastric cancer . However, an Austrian study, with an average of 8.4 years follow-up, reported associations between high fasting blood-glucose and a number of cancers, including: liver cancer and non-Hodgkin's lymphoma in men; and breast, colorectal and bladder-cancer in women . A UK case-control study showed increased HbA1c-levels among individuals with diabetes to be associated with pancreatic cancer incidence . A meta-analysis demonstrated associations of high levels of glucose and HbA1c with incidences of colorectal and pancreatic cancers . A more recent meta-analysis, including nine prospective studies, showed an association between fasting glucose and pancreatic cancer . A historical cohort study from a healthcare system in Wisconsin  reported no association between glycemic control and breast or colon cancer, but greater prostate cancer risk with better glycemic control (HbA1c≤7gr%).