Changes in dietary risk factors of colon cancer in Africans during urbanisation

Abstract

SUMMARY

Background The prevalence and incidence rates of colorectal cancer are significant lower in black South Africans than those in white, coloured or Indian South Africans. The reason for this phenomenon is not known, but it is suspected that blacks may have a genetic protection against development of this cancer and/or that environmental and lifestyle-related risk factors for colorectal cancer in blacks have not reached levels where the disease is "precipitated" in genetically susceptible individuals.

Objectives The objectives of this study were to review the literature on prevalence and incidence as well as risk factors for colorectal cancer, to measure the impact of urbanisation on changes in the diet-related risk factors in apparently healthy Africans, and to develop a hypothesis that may explain the low rates in Africans.

Setting The Northwest Province of South Africa and the Nutrition Laboratory of the Department of Nutrition and Family Ecology of the Potchefstroom University for Christian Higher Education.

Methods Study design and subjects: The study was a cross-sectional epidemiological survey of 889 apparently healthy African volunteers recruited from rural and urban parts of the Province. The study formed part of the TH USA project (Transition and Health during Urbanisation of South Africans) . The subjects, aged 15 to 65 years, were stratified into gender and 10-year interval age groups, as well as into 5 strata based on level of urbanisation: stratum 1 were subjects from deep rural areas; stratum 2 those living on farms; stratum 3 were living in squatter camps, and were mainly urban subjects which recently moved from rural areas; stratum 4 represented those urban subjects living in established townships while stratum 5 consisted of professional urban dwellers such as doctors, nurses and teachers .

Methodology Demographic and dietary information, as well as reproductive habits from women and bowel habits from selected subjects, were obtained during individual interview using validated and standardised questionnaires. Anthropometric measurements, clinical signs of malnutrition, oral temperature, blood pressure and two-hour glucose tolerance tests were performed by trained researchers and fieldworkers. Blood samples were collected and serum, citrated and EDTA plasma prepared within 3 hours and stored in aliquots, first at -20°C and later at -84°C , for later analyses of biochemical variables with standardised methods. Haematocrit and haemoglobin concentrations on EDTA blood were determined immediately, using a haematocrit centrifuge and capillary tubes and the method of Boehringer Mannheim respectively. All data were computerised and descriptive statistics, analyses of variance and Pearson correlation coefficients determined with the SAS® package by the Statistical Consultation Services of the PU vir CHO.

Results The results clearly illustrated an effect of urbanisation on accepted risk factors for chronic diseases of lifestyle. For example, progressive increases in diastolic blood pressure and total serum cholesterol were observed. The results further indicated that although reported energy intakes did not increase with urbanisation, the sources of energy changed. Total protein intakes were adequate but urban subjects ate more animal protein foods and less plant proteins than rural subjects, mainly because of an increase in meat and chicken intake. The exception was egg intake. Except for the farm workers who reported very low egg intakes, mean intakes of men were approximately 30g per day and women 20g/day. No effect of urbanisation was seen. The percentage of energy provided by fat progressively increased across strata while the percentage of energy derived from carbohydrates decreased. Subjects in stratum 5 already took in >30% of their energy as fat but still have relatively high carbohydrate intakes (>50% of energy) . Dietary fibre intakes were low in all strata (mean intakes from 10.9 to 24.2g/day) and remained low with urbanisation. Sucrose intakes were also low (30-36g/day) and did not increase with urbanisation. Mean starch intakes decreased across strata from 286 to 224g/day in men and 238 to 193g/day in women. Urban men and women also ate more meat, chicken, fruit and vegetables. This resulted in higher intakes of micronutrients such as vitamins E, C, folate and iron by urban subjects. The exception was higher calcium intakes by rural subjects (farm workers). Alcohol intakes were generally high in men. Dwellers of squatter camps reported the highest intakes. Patterns of reported intakes correlated with serum gamma glutamyl transferase values. In addition to these diet-related risk factors for colorectal cancer, other risk factors such as BMI and adult height increased with urbanisation. Physical activity seemed not to change with urbanisation, while smoking habit actually declined. However, the number of young people smoking was high.

Discussion Of the 13 accepted, probable or possible risk factors for colorectal cancer considered, eight in men and seven in women were related to low risk, while four in men and five in women were related to high risk. On the available information it was not possible to judge the effects or contribution of physical activity to risk. Of the factors associated with low risk, six: red meat, starch and fat (total and saturated) intake, BMI in women, as well as adult height, changed with urbanisation in a direction to increase risk. However, of these six, three (red meat and starch intake and adult height), although changed, were still under cut-point levels where they present increased risk. Of the five factors associated with an increased risk of colorectal cancer (low vegetable intake, smoking habit, high alcohol consumption by men, low dietary fibre intakes, and high BMI in women) three actually improved with urbanisation, possibly conferring a lower risk. These were increases in vegetable and fibre intake and a decrease in smoking habits of urban men, especially those in stratum 5. Because the contribution of a particular risk factor to total risk of colorectal cancer is not known, it is not possible to judge whether these changes will result in a lower overall risk. However, it is clear that with urbanisation, detrimental effects in some risk factors are not large enough to reach levels where they operate as risk factors, while in some instances, beneficial changes were observed in factors associated with high risk, possibly resulting in a lowering of risk.

Conclusion In conclusion it is hypothesised that at this stage of the lifestyle and nutrition transition that black South Africans are experiencing during urbanisation, the effects on the risk factors for colorectal cancer are such that some factors associated with high risk in rural subjects are improved while detrimental effects on some factors associated with a low risk are not large enough to "push" these factors over a level where they will be associated with a high risk. On balance it seems that at the present stage of urbanisation, changes in lifestyles and eating patterns generally are associated with lower risk of colorectal cancer. This is in contrast to some of the other chronic diseases of lifestyle such as hypertension and diabetes mellitus which already have emerged as a serious public health problem. It is possible that with further changes, especially further increase in meat and fat intake, the risk factor situation may change to such an extent that risk for colorectal cancer will increase. It is recommended that preventative measures such as promotion of a prudent, cancer-preventing diet should be targeted at black South Africans in transition.