Food intake and risk of cutaneous melanoma in an Italian population

Abstract

We investigated the association of melanoma risk with food consumption in a northern Italian population in which disease risk was shown to correlate with linoleic acid and soluble carbohydrates intake. We collected information regarding the habitual consumption of 188 food items in 59 patients with newly diagnosed cutaneous melanoma and 59 sex-and age-matched population controls. In the unadjusted analyses, the intake of several foodstuffs directly or inversely correlated with melanoma risk. In multivariate analysis adjusting for several potential confounders, risk correlated directly with vegetable oil intake and inversely with consumption of crispbreads and rusks. Overall, most of the food items rich in linoleic acid and soluble carbohydrates were unrelated to disease risk. Despite the limited statistical precision of the point estimates, these findings seem to indicate that consumption of specific foods may influence melanoma risk. IntroductionThe marked differences in melanoma incidence over time and between different populations suggest a key etiologic role of environmental factors, diet being among these potential risk factors (Millen et al., 2004;Naldi et al., 2004;Granger et al., 2006;Le Marchand et al., 2006). We observed recently in a northern Italian population an association between melanoma risk and a high intake of linoleic acid and a low consumption of soluble carbohydrates, and we aimed at identifying in the same population the food items specifically affecting disease risk (Vinceti et al., 2005a). MethodsWe carried out a case-control study to investigate the role of environmental risk factors in the etiology of cutaneous melanoma in Modena province, northern Italy (around 6 50 000 inhabitants) (Vinceti et al., 2005a, b). We examined a consecutive series of 59 newly diagnosed melanoma patients attending the Dermatologic Clinic of Modena University Hospital, provided that they resided in province and had recently been diagnosed with melanoma at an early clinical stage. Each patient was associated with a control through random selection among age-(75 years) and sexmatched Modena province residents.Apart from a dermatological examination, collection of anamnestic data and sampling of blood and toenails specimens, we obtained information from cases and controls regarding their normal diet through a self-administered semiquantitative food frequency questionnaire originally developed and validated within the European Prospective Investigation into Cancer and Nutrition (EPIC) program (Pisani et al., 1997(Pisani et al., , 2002Pala et al., 2003). This questionnaire was specifically designed for the central northern Italy and consisted of 248 questions concerning the frequency and the quantity of consumption of 188 items, and it was designed to ascertain in detail how much and what kind of foods had been consumed over the previous year, using illustrations of sample dishes of definite sizes or by reference to standard portion sizes.We calculated melanoma relative risk (RR), as estimated through the odds ...

Systematic review of studies investigating the association between dietary habits and cutaneous malignant melanoma

“…Seven studies [37,40,41,44,45,48,53] dealt with fish consumption, nine [37, 40, 41, 44-47, 53, 54] with vegetables and eight [37,40,41,[44][45][46][47][48] with fruit. Most studies did not identify any significant association between foods intake and CMM, except for two studies [37,45] out of seven (28.6%) demonstrating a significant reduction of about 35-37% associated to a higher fish consumption, three [37,45,47] out of nine (33.3%) showing a 40-57% risk reduction with a higher vegetables intake and two [37,47] out of eight (25%) highlighting a significant reduction ranging from 34% to 46% with respect to fruit.…”

“…In conclusion, sixteen articles [36, 37, 40-42, 44-49, 51, 53-54, 56, 61] were considered in the review (Figure 1) [40,42] were considered as a single study, because the sample was exactly the same.…”

“…Every paper considered the comparability of cases and controls controlling the analysis for the most important confounding factors. About the exposure, the ascertainment was given by secure records or structured interview blind to case/control status in four studies [37,40,42,44,54] but methods of ascertainment were the same for cases and controls in all papers.…”

“…Seven studies (46.7%) [36,37,40,42,48,54,61] out of fifteen were conducted in Europe, mostly Italy; only one (6.6%) was conducted in Australia [56]; the remaining seven (46.7%) were carried out in United States [41, 45-47, 49, 51, 53].…”

“…In particular, as far as the type of questionnaire is concerned, studies were performed as follows: a semi-quantitative food-frequency questionnaire was administered in five studies [36,40,42,48,51,53]; in two studies a food-frequency questionnaire was used [46,49]; in only one study the "Dobson short fat questionnaire" was used [56]; in four studies trained interviewers administered a structured questionnaire [37,44,54,61]; the study subjects were requested to complete a reduced version of the "Block-food-frequency questionnaire" in another study [45]; in only one study each participant completed a telephone survey and a questionnaire was administered during the phone interview [47]; last, in only one study in-persons interviews were conducted at subject's homes by trained interviewers [41]; -methods used in order to quantify servings were both qualitative (e.g. pictures or images) and quantitative (e.g.…”