In a previous report,1 we provided information on the relation between acrylamide intake and risk of cancers of the upper aero-digestive tract, colorectum, breast, ovary and prostate, using data from a series of Italian and Swiss case-control studies. No significant association emerged for any of the cancer sites considered, with odds ratios (OR) for the highest level of acrylamide intake ranging between 0.92 for prostate cancer and 1.23 for laryngeal cancer. New data are now available on renal cell cancer (RCC) from an Italian case-control study conducted using the same methods and food frequency questionnaire. To our knowledge, only 2 Swedish case-control studies have provided data on the issue,2, 3 and found no association between acrylamide intake (OR = 0.8; 95% confidence interval (CI), 0.4–1.7, and OR = 1.1; 95% CI, 0.7–1.8, for the highest versus lowest quartile) and risk of kidney and RCC, respectively.
The general scheme of the study and methods have been described elsewhere.1, 4 Briefly, a case-control study of RCC was conducted between 1992 and 2004 in 4 Italian areas. Cases were 767 patients (494 men, 273 women) with incident, histologically confirmed RCC, recruited in a network of major teaching and general hospitals in the areas under surveillance. Controls were 1,534 subjects (988 men, 546 women), matched with cases by study centre, sex and age, admitted to the same hospitals for acute non-neoplastic conditions, unrelated to known or potential risk factors for RCC nor to long-term modifications of diet.
Cases and controls were questioned during their hospital stay by trained interviewers. Information on diet was based on a food frequency questionnaire including 78 foods and recipes, that showed satisfactory validity and reproducibility.5, 6 Among the items in the food frequency questionnaire, those considered for the estimation of acrylamide intake were consumption of fried/baked potatoes, coffee, “cappuccino” (white coffee), bread, pizza, fried meat and fish, sweet biscuits, crackers, breadsticks and melba toast. The mean acrylamide content of each food was obtained from resources made available on the Internet by the World Health Organization (WHO, http://www.who.int/foodsafety/chem/en/) and the French Agency for Food Safety (AFSSA, http://www.afssa.fr).7, 8 For those foods that are prepared in a similar way in Italy and France (e.g., bread), we preferred - when available - data from the latter source.
We estimated the ORs and the corresponding 95% CI using multiple conditional logistic regression models, conditioned on study centre, sex and quinquennia of age, and adjusted for year of interview, education (<7, 7–11, ≥12 years), smoking habit (never, ex, current smokers of <15, 15 to < 25, ≥25 cigarettes per day), alcohol consumption (never, ex, current drinkers of <21, ≥21 drinks per week), body mass index (<25, 25 to < 30, ≥30 kg/m2), occupational physical activity (mainly standing/sitting, intermediate, heavy/strenuous), family history of kidney cancer and energy intake (in quintiles).
Table I shows the frequency distribution of cases and controls, ORs and 95% CI according to consumption of selected foods containing acrylamide and total acrylamide intake. Fried/baked potatoes was the food item that contributed most to total dietary acrylamide (29.6%), and its consumption was not related with RCC (OR = 1.06, 95% CI 0.78–1.43, for >1 portion/week as compared to no consumption). White bread contributed for 28.6% of total acrylamide intake, and the corresponding OR of RCC for high consumption was 1.70 (95% CI, 1.25–2.30). Sweet biscuits (15.0% of total dietary acrylamide, OR = 0.93 for high consumption), coffee (12.4%, OR = 1.03), crackers (6.5%, OR = 1.07) and other food items that contributed to a minor proportion of total acrylamide intake were not associated with RCC risk.
|Foods and total acrylamide intake||Cases:controls||OR (95% CI)1||Total acrylamide2 (%)||Mean intake ± SEM3|
|Fried/baked potatoes (portions/week)||29.6||0.63 ± 0.02||0.66 ± 0.02|
|White bread (portions/week)||28.6||18.00 ± 0.48||16.46 ± 0.33|
|7 to <21||402:752||1.49 (1.18–1.87)|
|Sweet biscuits (portions/week)||15.0||2.27 ± 0.17||2.07 ± 0.11|
|Coffee (portions/week)||12.4||17.25 ± 0.47||16.72 ± 0.29|
|7 to <21||377:710||1.17 (0.90–1.54)|
|Crackers (portions/week)||6.5||3.00 ± 0.18||3.00 ± 0.15|
|Total acrylamide intake (μg/day)||–||–||40.7 ± 1.36||37.1 ± 0.95|
|1st quartile (<20.4)||172:384||1|
|2nd quartile (20.4–31.2)||203:383||1.21 (0.94–1.57)|
|3rd quartile (31.2–44.1)||192:383||1.14 (0.86–1.51)|
|4th quartile (>44.1)||200:384||1.20 (0.88–1.63)|
|p for trend||0.35|
|Continuous OR4||1.05 (0.94–1.16)|
Our estimate of average acrylamide intake (37 μg/kg, 0.475 μg/kg body weight/day) was consistent with published data from other populations.9 The ORs for increasing quartiles of total acrylamide intake were 1.21 (95% CI, 0.94–1.57), 1.14 (95% CI, 0.86–1.51) and 1.20 (95% CI, 0.88–1.63) as compared to the lowest quartile, with no trend in risk (p = 0.35). When we excluded the contribution of white bread consumption from the estimate of acrylamide intake, the OR for the highest vs. lowest quartile of acrylamide was 0.87 (95% CI, 0.65–1.14). Risk of RCC according to total acrylamide intake was consistent across strata of sex and age (data not shown).
With reference to potential bias of this study, there is no reason to assume a different recall of foods containing acrylamide between cases and controls, because most of these foods are not known to be related with cancer risk. Further, the similar interview setting of cases and controls provides reassurance against information bias. The almost complete participation of cases and controls and the inclusion of acute conditions unrelated to diet in the comparison group weigh against any major role of selection bias. Moreover, the main dietary results of this study are in broad agreement with data from previous investigations.4 Another potential source of bias is misclassification of acrylamide intake. In a sensitivity analysis, we used median acrylamide content instead of mean values from the French database. Although estimated intakes were about 10% lower, none of the results was materially different.
Therefore, this study confirms the absence of significant relations between dietary acrylamide and risk of RCC, in line with the findings of 2 Swedish case-control studies.2, 3 These also reported no association between food items containing elevated levels of acrylamide and risk of kidney and RCC. On the other hand, the modest increase in risk observed in our study for the 3 upper quartiles of acrylamide intake compared to the lowest one is likely attributable to the direct relation found between white bread consumption and RCC. This might be explained by the high glycemic index of this food and its consequent effect on insulin-like growth factors levels,10 as was discussed in an earlier report from this study.4