Heating Cruciferous Vegetables Increases in Vitro Dialyzability of Intrinsic and Extrinsic Iron ACHILLEFS N. KAPANIDIS and TUNG-CHING LEE ABSTRACT Iron dialyzability (ID) from three Cruciferae (broccoli, kale and cab- bage) was determined by an in vitro digestion method. The effect of added Crucifers on the ID of extrinsic nonheme iron as well as effects of heating were studied. Uncooked Crucifers contained iron of moderate ID (7-9%); cooking resulted in substantial ID increase (200%). Cooked Crucifers increasedextrinsic ID three- to fourfold. Time and temperature relationships for the increase suggestedthat organic acids released after partial cell wall degradation combined with protein denaturation and iron solubilization from fiber were major reasonsfor differences between raw and cooked vegetables. Cruciferous vegetables can contribute to im- proved iron nutrition. Key Words: iron, in vitro dialyzability, bioavailability, Crucijbae, fiber INTRODUCTION IRON DEFICIENCY ANEMIA, the most prevalent nutritional prob- lem affects 500-600 million people worldwide (Baynes and Bothwell, 1990). Diet composition is critical since low iron in- take and/or bioavailability (IB) are main causes of the deficiency (Fairweather-Tait, 1984). Iron contributes to two pools in the body, for heme iron (HI) and nonheme iron (NHI) which ac- count for 10 and 90% of dietary iron respectively. HI absorption is high and diet-independent (Sweeten et al., 1986); NH1 ab- sorption, although variable, is considered low and greatly influ- enced by interactions with enhancers and inhibitors (Layrisse et al., 1969; Cook, 1983). Meat (Monsen et al., 1978), ascorbate (Monsen, 1982) and citrate (Gillooly et al., 1983) increases IB whereas fiber (Reinhold, 1982), tannins (Disler et al., 1975), phytate (Ton-e et al., 1991) and certain proteins (Kratzer and Vohra, 1986) act as inhibitors. Iron from different foods exhibits variable IB. Meat is a sig- nificant iron source as it contains bioavailable HI. However, NH1 found in cereals, nuts and legumes exhibits low IB due to high concentrations of inhibitors (Hazel1 and Johnson, 1986). High variability has been observed among vegetables (Gillooly et al., 1983). Some, such as spinach (Gordon and Chao, 1984), provide poor sources of iron and only speculations for signifi- cant IB exist for others. The Cruciferae family includes vege- tables such as broccoli, cabbage and several green leafy vegetables. They contain high amounts of important nutrients and anti-cancer agents which have brought about increased in- terest in them (Caragay, 1992). The presence of iron and organic acids in Crucifers indicates they may serve both as a source of bioavailable NH1 and an enhancer of NH1 absorption from other foods. Sharp IB changes may occur after processing, since it affects the food matrix and alters the iron chemistry (Lee, 1982) making processing optimization necessary for iron retention and IB. No systematic in vitro study of IB has been reported on Crucifers. We used the Miller et al. (1981) method to estimate in vitro iron dialyzability (ID). Schricker et al. (1981) have shown that this fast and low cost method correlated highly with in vivo data on identical meals and that ID could provide a reliable index of Authors Kapanidis and Lee are affiliated with the Dept. of Food Science, and the Center of Advanced Food Technology, Rutgers, the State Univ. of New Jersev, New Brunswick, NJ 08903-0231. IB. Our objectives were (a) to estimate the in vitro ID of three raw and cooked Crucifers, (b) to study the effects of their ad- dition on the in vitro ID of extrinsic iron, and (c) to evaluate the effects of cooking on the ID of Crucifers. MATERIALS & METHODS Glassware and reagents All glassware used was washed with distilled water, left overnight in 1M sulfuric acid bath to remove possible iron contamination and rinsed several times with distilled deionized water. Pepsin solution. 16g of pepsin 1:10,000 powder from hog stomach (Sigma) were dissolved in 100 mL 0.1 M HCl (trace metal grade, Fisher). The solution was prepared fresh weekly and kept at 4°C. Pancreatin-bile (PB) suspension. 1 g pancreatin from porcine pan- creas and 6.5g porcine bile extract (both from Sigma) were suspended in 250 mL of 0.1 M NaHCO,. The suspensionwas prepared fresh weekly and stored at 4°C. Iron stock solution. 1015 ppm Fe3+ stock solution (Aldrich, Milwau- kee, WI) in O.lN HCI was used for preparation of AAS standard solu- tions. Preparation of vegetables Broccoli (Brussica oleoracea L. var. italica Plenk) kale (Brassica oleoracea L. var. acephala DC.) and cabbage (Brassica oleoracea L. var. capitata) were studied (average 1.3, 1.9 and 0.5 ppm Fe respec- tively). The vegetables were purchased from a local supermarket (New Brunswick, NJ) and analyzed immediately or after brief storage. We washed, air-dried and cut the vegetables, removing inedible parts. We cooked the vegetables by boiling cut leaves or florets in a covered glass beaker containing 100-200 mL distilled deionized watei (depending on water content of vegetable). Use of glass exclused iron leaching expe- rienced with metallic utensils (Mistry et al., 1988). Boiling time was 15 min, measured after adding the vegetables in boiling water. At the end of treatment, the beaker and its contents were quickly cooled to room temperature (-25°C) using running tap water (to avoid changes due to elevated temperature after 15 min). We retained the boiling water as it contained appreciable soluble (and thus highly bioavailable) iron (Gil- looly et al., 1983; Latunde-Dada, 1990) and used it for preparation of initial homogenates. Microwave cooking was performed in a 600 W microwave oven with rotating base (81-MDS, CEM Corp.). Broccoli was heated for various times in a covered glass beaker containing 50 mL water. ID values of raw and cooked vegetableswere determined the same day. The ID values for each vegetable are averagesof triplicate batches with three values for each batch. Iron determination in raw vegetables We dried the vegetables (AOAC, 1990) to avoid errors caused by water evaporation. Dried samples were ground and 0.5 g was weighed in Teflon digestion vessels (CEM Corp.). We added 10 mL N HNO, and after 30 min, we added 5 mL of 10 N HCl. After gas production subsided, we capped and placed the vessels in the microwave digestion oven turntable (MDS-81D, CEM Corp.) The samples were digested us- ing 100% power for 5 min and 70% for 8 min (internal pressure mon- itored by CEM pressure controller). We diluted the digests to 25 mL, filtered and transferred them in vials. Iron concentration of the digests was determined by Atomic Absorption Spectrophotometer(AAS, model 603, Perkin-Elmer Corp.) with air-acetylene flame at 248.3 nm (slit length 2 nm). The calibration curve was established using standards of 0. I, 0.3, 0.5, 1, 2, 4, 5, 8 and IO ppm Fe. The relationship between iron concentration and absorbance was linear with high correlation coeffi- cients (RQ0.99). 128-JOURNAL OF FOOD SCIENCE-Volume 60, No. 1, 1995