Food groups and renal cell carcinoma: A case–control study from Italy
Article first published online: 20 OCT 2006
Copyright © 2006 Wiley-Liss, Inc.
International Journal of Cancer
Volume 120, Issue 3, pages 681–685, 1 February 2007
How to Cite
Bravi, F., Bosetti, C., Scotti, L., Talamini, R., Montella, M., Ramazzotti, V., Negri, E., Franceschi, S. and La Vecchia, C. (2007), Food groups and renal cell carcinoma: A case–control study from Italy. Int. J. Cancer, 120: 681–685. doi: 10.1002/ijc.22225
- Issue published online: 29 NOV 2006
- Article first published online: 20 OCT 2006
- Manuscript Accepted: 22 JUN 2006
- Manuscript Received: 9 MAY 2006
- renal cell carcinoma;
- case–control study;
- risk factors
Although nutrition and diet have been related to renal cell carcinoma (RCC), the role of specific foods or nutrients on this cancer is still controversial. We evaluated the relation between a wide range of foods and the risk of RCC in an Italian case–control study including 767 patients (494 men and 273 women) younger than 79 years with incident, histologically confirmed RCC, and 1,534 controls (988 men and 546 women) admitted to the same hospitals as cases for a wide spectrum of acute, non-neoplastic conditions, not related to long term diet modifications. A validated and reproducible food frequency questionnaire, including 78 foods and beverages, plus a separate section on alcohol drinking, was used to assess patients' dietary habits 2 years before diagnosis or hospital admission. Multivariate odds ratios (OR) were obtained after allowance for energy intake and other major confounding factors. A significant direct trend in risk was found for bread (OR = 1.94 for the highest versus the lowest intake quintile), and a modest excess of risk was observed for pasta and rice (OR = 1.29), and milk and yoghurt (OR = 1.27). Poultry (OR = 0.74), processed meat (OR = 0.64) and vegetables (OR = 0.65) were inversely associated with RCC risk. No relation was found for coffee and tea, soups, eggs, red meat, fish, cheese, pulses, potatoes, fruits, desserts and sugars. The results of this study provide further indications on dietary correlates of RCC, and in particular indicate that a diet rich in refined cereals and poor in vegetables may have an unfavorable role on RCC. © 2006 Wiley-Liss, Inc.
Renal cell carcinoma (RCC), the major histologic type of kidney cancer, accounts for about 2% of all cancers in adults, in many western countries.1, 2 Over the last few decades, its incidence has been rising, though this may be partially due to the increment of incidental diagnoses resulting from the widespread use of abdominal imaging procedures.3 Several potential risk factors have been considered in previous epidemiological studies, but apart from a role of tobacco smoking,4, 5 obesity,2, 4, 6 family history of kidney cancer,7, 8, 9 hypertension,4 and phenacetin use,4 the etiology of this cancer remains largely undefined.
Diet has been related to RCC, although the role of specific foods or nutrients is still controversial. Several studies reported inverse relations with vegetables1, 4, 10, 11, 12, 13 and fruits,1, 4, 10, 11, 12, 13 and with specific items such as dark-green,14, 15, 16 cruciferous15, 16 and root vegetables,13, 17 carrots18 and citrus fruits.12 However, other investigations did not confirm these associations.17, 18, 19, 20, 21 Foods rich in proteins have been directly related to RCC,4, 11, 22 but the findings are still controversial,8, 23 and appear limited to specific food items, such as red and processed meat,16 fried meat12, 14 or poultry. Other relations have been sporadically considered with milk and dairy foods,4, 10 coffee1, 10 and tea drinking,10 but the role of these foods on the risk of RCC remains inconclusive.8
We have, therefore, analyzed the role of a wide range of foods and foods groups on the risk of RCC, using data from a large case–control study conducted in Italy.
Material and methods
A case–control study on RCC was conducted between 1992 and 2004 in four Italian areas, including the greater Milan area and the provinces of Udine and Pordenone in northern Italy, the province of Latina in central Italy and the urban area of Naples in southern Italy. Cases were 767 patients (494 men and 273 women) under age 79 years (median age 62 years, range 24–79 years) with incident, histologically confirmed RCC (ICD-9 189.0), admitted to major teaching and general hospitals of the study areas. Cancers of the renal pelvis and ureter (ICD-9 189.1–189.2) were not included. Controls were 1,534 subjects (988 men and 546 women) under age 79 years (median age 62 years, range 22–79 years) admitted to the same hospitals as cases for a wide spectrum of acute nonneoplastic conditions, unrelated to known or potential risk factors for RCC. Controls were matched with cases by study center, sex and quinquennia of age, with a case to control ratio of 1:2. Twenty-six percent of controls were admitted for traumas (mostly fractures and sprains), 32% for other orthopedic disorders (such as low back pain and disc disorders), 14% for surgical conditions and 27% for various other illnesses including eye, nose, ear, skin or dental disorders. Chronic conditions, likely related to long-term diet modifications, were excluded. Less than 5% of both cases and controls contacted refused to participate.
Cases and controls were interviewed during their hospital stay by trained interviewers, using a structured questionnaire, including information on sociodemographic characteristics, anthropometric measures, lifestyle habits, such as tobacco smoking, alcohol and coffee consumption, personal medical history and family history of cancer in first-degree relatives.
The subjects' usual diet during the 2 years before cancer diagnosis or hospital admission (for controls) was assessed using an interviewer-administrated FFQ including 78 foods and beverages, as well as a range of the most common Italian recipes, grouped into 6 sections: milk and hot beverages; bread and cereals dishes (first course); meat and other main dishes (second course); vegetables (side dishes); fruits; sweets, desserts and soft drinks; plus an additional section on alcoholic beverages. Subjects were asked to indicate the average weekly frequency of consumption for each dietary item; intakes lower than once a week but at least once a month were coded as 0.5 per week. A commonly used unit or serving size was prespecified for 40 items; for the remainder, the portion was defined as small, average or large, with the help of figures. For a few vegetables and fruits, seasonal consumption and the corresponding duration was elicited. To estimate the total energy intake, an Italian food composition database was used, integrated with other sources when needed.24, 25 The FFQ was satisfactorily valid26 and reproducible,27 with Spearman correlation coefficient between 0.60 and 0.80 for most food items.
Food items were combined in 21 food groups: milk and yoghurt; coffee and tea; bread; pasta and rice; soups; eggs; poultry; red meat; processed meat; fish; cheese; pulses (i.e., green peas, beans, lentils, etc.); raw vegetables; cooked vegetables; all vegetables; potatoes; citrus fruits; other fruits; all fruits; desserts; sugars. The weekly intake for each group was calculated summing up the intake of each food item included in the group, and was then distributed into approximate quintiles among controls (quartiles or tertiles for a few food groups not frequently consumed).
Odds ratios (OR) and the corresponding 95% confidence intervals (CI) were calculated using conditional multiple logistic regression models,28 conditioned on study centre, sex and quinquennia of age and adjusted for period of interview, years of education, family history of kidney cancer in first-degree relatives, tobacco smoking, alcohol drinking, body mass index (BMI) and total energy intake. Tests for trend were based on the likelihood ratio test between models with and without a linear term for each food group. The OR for an increase of one average serving per day was also estimated, including each food item in the model as a continuous variable.
Table I shows the distribution of 767 cases of RCC and 1,534 controls, according to selected covariates. By design, cases and controls had the same sex and age distribution. Cases were more educated than controls, were more frequently heavy smokers, had a higher BMI and reported more often first-degree relatives with kidney cancer. As regards alcohol drinking and total energy intake, cases and controls were comparable.
|Body mass index (kg/m2)1|
|Family history of kidney cancer4|
|Total energy intake (kcal/day)|
Table II gives the multivariate ORs of RCC for intake quintiles of various food groups. A significant direct association was observed for bread consumption (OR = 1.94, 95% CI: 1.40–2.71, for the highest compared to the lowest quintile of intake, p for trend = 0.0002). A modest excess of risk was also observed for pasta and rice (OR = 1.29, 95% CI: 0.95–1.76, p for trend = 0.06), and milk and yoghurt (OR = 1.27, 95% CI: 0.95–1.70, p for trend = 0.11). Conversely, a significant trend of decreasing risk was observed with increasing intake of poultry (OR = 0.74, 95% CI: 0.58–0.94, p for trend = 0.008), processed meat (OR = 0.64, 95% CI: 0.45–0.90, p for trend = 0.006), raw vegetables (OR = 0.74, 95% CI: 0.53–1.02, p for trend = 0.007), cooked vegetables (OR = 0.68, 95% CI: 0.50–0.91, p for trend = 0.01) and all vegetables (OR = 0.65, 95% CI: 0.47–0.90, p for trend = 0.003). Intake of coffee and tea, soups, eggs, red meat, fish, cheese, pulses, potatoes, citrus fruits, other fruits, desserts and sugars were not significantly related with RCC.
|Food group (servings per week)||Quintile of intake, OR1 (95% CI)||χ2 trend (p value)|
|Milk and yoghurt|
|Quintile upper limit||0.29||3.2||7.2||9.0||–||2.61 (0.11)|
|Coffee and tea|
|Quintile upper limit||7.4||13.9||20.9||27.9||–||0.002 (0.96)|
|Quintile upper limit||9.5||14.75||21.25||28.5||13.56 (0.0002)|
|Pasta and rice|
|Quintile upper limit||3.25||4.25||5.25||6.5||3.63 (0.06)|
|Quintile upper limit||1.9||2.9||–||1.13 (0.29)|
|Quintile upper limit||0.9||1.9||–||1.42 (0.23)|
|Quintile upper limit||0.9||1.9||–||7.07 (0.008)|
|Quintile upper limit||2.4||3.4||4.4||5.9||–||1.92 (0.17)|
|Quintile upper limit||0.9||1.9||2.9||3.9||–||7.53 (0.006)|
|Quintile upper limit||0.9||1.9||–||3.17 (0.07)|
|Quintile upper limit||2.2||3.3||4.3||5.9||–||0.36 (0.55)|
|Quintile upper limit||0.4||0.9||–||1.58 (0.21)|
|Quintile upper limit||3.4||4.9||8.4||11.9||–||7.29 (0.007)|
|Quintile upper limit||1.7||2.7||3.6||4.8||–||5.96 (0.01)|
|Quintile upper limit||6.2||9.2||12.4||15.7||–||8.77 (0.003)|
|Quintile upper limit||0.9||1.9||–||0.15 (0.70)|
|Quintile upper limit||0.9||2.4||3.4||6.9||–||0.83 (0.36)|
|Quintile upper limit||6.2||9.9||13.4||18.4||–||0.001 (0.98)|
|Quintile upper limit||6.4||10.2||13.6||18.4||–||0.0003 (0.99)|
|Quintile upper limit||0.6||1.9||4.2||8.1||–||1.64 (0.20)|
|Quintile upper limit||14.9||27.9||41.9||64.4||–||0.39 (0.54)|
Table III shows the ORs of RCC for an increase of consumption of one serving per day for selected food groups which showed a significant trend in risk, in strata of sex, age (<60, ≥60 years), tobacco smoking (never, ever smokers) and BMI (<25, ≥25 kg/m2). The associations between the food groups considered and RCC were consistent across strata of these covariates, the estimates being non significantly heterogeneous.
|Food group||OR1 (95% CI)|
|Overall||Sex||Age (years)||Smoking habit||Body mass index (kg/m2)|
|Men||Women||<60||≥60||Never smokers||Ever smokers||<25||≥25|
|Pasta and rice||1.43(1.04–1.96)||1.36(0.93–2.00)||1.47(0.82–2.62)||1.42(0.89–2.27)||1.45(0.94–2.25)||1.48(0.87–2.54)||1.38(0.92–2.07)||1.68(0.95–2.97)||1.25(0.85–1.86)|
The results of the present study—based on one of the largest case–control investigations on RCC—provide several interesting indications on the dietary correlates of this cancer in a southern European population. A direct association was found for cereals, mainly bread, while inverse relations were reported for vegetables and various types of meat.
With reference to cereals, increased risks were found for increasing consumption of both bread, and, to a lesser extent, of pasta and rice. No significant relation was found for potatoes, another important source of starch, although their risk estimates were compatible with a modest increased risk. To our knowledge, no other study investigated the role of cereals on RCC. As for other common cancers, the increased risk of RCC for elevated cereals intake may be due to the high glycemic index of these foods, and their possible involvement in insulin-like growth factors.29
The inverse relation between vegetable consumption and RCC observed in the present study is consistent with the results of several previous studies1, 4, 10, 12, 13, 16, 30 Further, a protection was found for both raw and cooked vegetables, thus suggesting that any possible favorable effect in vegetables is not adversely affected by the cooking process. The beneficial effect of vegetables may be related to their content of vitamins, micronutrients or food constituents, such as carotenoids, flavonoids and phytosterols, known to have antioxidant and anticancerogenic proprieties.31 However, the components of vegetables which may have a favorable role on RCC remain undefined.
No association was found between fruit consumption and RCC risk, even considering different types of fruit. Similarly, 2 previous cohort studies found no association between RCC risk and fruit consumption,17, 21 though a few other investigations reported a beneficial effect.1, 4, 10, 11, 12, 13
An inverse relation was observed with poultry and processed meat consumption, while red meat was unrelated to RCC risk. With reference to other foods of animal origin, eggs and fish were not associated to RCC risk in the present study. A few previous studies suggested a possible association between RCC risk and meat, reporting increased risk for red meat,4, 16, 22 and poultry12 and generally for a dietary pattern rich in various types of meat,32 though the evidence of an effect of meat on RCC risk remains inconclusive.8, 23
The apparent inverse relation between various sources of proteins on RCC observed in this study may be due to selective reduction of protein-rich diets in subjects with kidney or urinary tract diseases. An association between these diseases and RCC has been reported in several—though not all—case–control studies,8 but may be partially or largely due to information bias, i.e., more accurate reporting of renal diseases by kidney cancer patients. Thus, the issue of meat and other sources of protein on RCC remains unclear.
Regarding dairy products consumption, a modest increased risk of RCC was found for milk and yoghurt, but not cheese consumption. Thus, our results do not support—although they are not incompatible—the results of a few studies4, 10 suggesting a detrimental role of dairy products on RCC.
We investigated the role of diet on RCC considering individual foods or food groups, rather than dietary patterns, which are more suitable to investigate the concurrent role of various foods on the disease. However, this kind of analysis is more complex and requires subjective choices to define the patterns; further its results depends on geographical and cultural variation of the dietary habits, and are difficult to interpret and generalize. Two studies have examined the role on the risk of RCC of dietary patterns identified through factor analysis, finding inconclusive results.30, 32
Dietary recall may be influenced by recent diagnosis of cancer as cases may recall their dietary habits differently from controls. However, this criticism applies less to hospital-based studies, as the controls are also ill and hospitalized. Further, it is unlikely that neither cases nor controls had any perception of a relation between dietary habits and RCC risk. The almost complete participation of both cases and controls, and the inclusion of acute conditions unrelated to diet in the comparison group, weighs against any major role of selection bias in our study. Interviewers were not blind to the case–control status, but the questionnaire was administered to both cases and controls by the same interviewers, under similar conditions. Other strengths of the present study are the comparable catchments areas of the study subjects, the satisfactorily reproducibility and validity of diet information26, 27, the wide sample size and the possibility to adjust for energy intake and other major confounding factors.
Our results confirm that diet may play a role on the risk of RCC, and in particular, a moderate cereal and high vegetables consumption may have a favorable effect on this neoplasm.
This work was conducted with the contribution of the Italian Association for Cancer Research and the Italian League Against Cancer. The authors thank Ms. M.P. Bonifacino for editorial assistance.
- 4Renal cancer. In: SchottenfeldD, FraumeniJF,Jr, eds. Cancer epidemiology and prevention, 2nd ed. New York: Oxford University Press, 1996. 1142–55., , , .
- 5IARC. Tobacco smoke and involuntary smoking. IARC monographs on the evaluation of carcinogenic risks to humans, vol. 83. Lyon, France: IARC, 2004.
- 10World Cancer Research Fund and American Institute of Cancer Research. Food, nutrition and the prevention of cancer: a global perspective. Washington, DC: World Cancer Research Fund, 1997.
- 24Banca di composizione degli alimenti per studi epidemiologici in Italia. Milano, Italia: Istituto Europeo di Oncologia, 1998., , , , .
- 28Statistical methods in cancer research, vol. 1. The analysis of case-control studies. IARC Sci Publ No. 32. Lyon, France: IARC, 1980., .