1120 Am J C/in Nuir 1992;55:h 120-5. Printed in USA. 1992 American Society for Clinical Nutrition

Plasma carotenoid response to chronic intake of selectedfoods and 3-carotene supplements in ,2Marc S Micozzi, Ellen D Brown, Brenda K Edwards, JGFred Khachik, Gary R Beecher, and J Cecil Smith, Jr

ABSTRACI We determined serial changes in four majorplasma carotenoid fractions (a-carotene, /3-carotene, lutein/zeaxanthin, and lycopene) in 30 men consuming defined dailydoses of carotenoids from foods (broccoli, carrots, or tomatojuice) or from purified /3-carotene in capsules (12 or 30 mg) for6 wk while fed a controlled diet. Compared with baseline, /3-carotene increased in the 30- and 12-mg-capsule and carrotgroups whereas a-carotene increased in the carrot group andlutein increased in the broccoli group. Lower lutein concentra-tions in recipients of/3-carotene capsules suggested an interactionbetween these two carotenoids. Lycopene declined in all groupsexcept the tomato-juice group. Total carotenoid concentrationchanges only reflected the large increases in /3-carotene concen-trations and not the smaller changes observed in other individualcarotenoids. Overall, purified /3-carotene produced a greaterplasma response than did similar quantities ofcarotenoids fromfoods sources. However, some foods increased plasma concen-trations of certain carotenoids. Am J C/in Nutr l992;55:1 120-5.

KEY WORDS /3-carotene, carotenoids, foods, lycopene, lu-tein, a-carotene, supplements

Introduction

Epidemiologic studies on diet and cancer have indicated thatconsumption ofcarotenoid-containing foods may be associatedwith a reduced risk of certain epithelial cancers in humans (1-4), although the association of such foods (eg, cabbage, carrots)with prevention and/or treatment of cancer has been discussedsince antiquity, beginning with the Roman, Cato the Elder (234-149 BC) (5). However, current attention to these foods has not

been directed by history but has been a result of epidemiologicassociations (6). Thus, the majority of chemotherapeutic andchemopreventive agents today are tested for efficacy in controlledclinical trials. Accordingly, numerous clinical trials are currentlyin progress to test whether chronic supplementation of purified/3-carotene may prospectively reduce cancer rates. It is importantto examine the plasma carotenoid response to /3-carotene sup-plementation and to compare this response with that obtainedwith dietary intake of carotenoid-containing foods, especiallythose that have been associated with reduced risk of cancer (7).

Therefore, we determined the serial changes in seven plasmacarotenoid fractions in 30 men given defined daily doses of ca-rotenoids from carotenoid-containing foods (broccoli, carrots,

Bieri, Philip R Taylor,

tomato juice) or /3-carotene from purified supplements ( 12 or30 mg) for 6 wk on a controlled diet.

Subjects and methods

Thirty healthy men participated in a chronic carotenoid sup-plementation study conducted during October-December 1985.Volunteers were screened by using medical and dietary ques-tionnaires, physical examination, hematologic and biochemicalprofiles, plasma carotenoid and retinol concentrations, and se-hected anthropometric measurements. Eligible subjects were be-tween the ages of 20 and 45 y, had no history ofchronic diseaseand no abnormalities on the basis of physical exams or bloodprofiles, were not smokers, were within 10% of ideal body weightfor size, had baseline plasma total carotenoid concentrationsbetween 1 .3 and 3.7 zmol/L, and had no particular dietary pat-terns (eg, vegetarianism). Individuals not conforming to theseeligibility criteria were excluded. Self-supplementation with vi-tamins was stopped by all subjects taking them 4 wk beforethe start ofthe study. Procedures used in the study were approvedby committees for human subjects protection from the UnitedStates Department of Agriculture, the National Cancer Institute,and Georgetown University. Subjects were separated into twostrata according to baseline total plasma carotenoid concentra-tions to identify subjects with concentrations above and belowthe mean. They were then randomly assigned to one ofsix treat-ment groups (five subjects per group) that were defined by theamount and type of total carotenoid supplementation received(Table 1).

Food analyses

All foods used as treatments were purchased in single lotsbefore the study began in amounts sufficient to last for the entiretreatment interval. Food samples were analyzed in the state of

I From the National Museum ofHealth and Medicine, Armed ForcesInstitute of Pathology, Washington, DC; the Division ofCancer Preven-tion and Control, NCI, Bethesda, MD; and the Beltsville Human Nu-trition Research Center, USDA, Beltsvilhe, MD.

2 Address correspondence and reprint requests to PR Taylor, CancerPrevention Studies Branch, National Cancer Institute, Executive PlazaNorth, Suite 2 1 1, Bethesda, MD 20892.

Receivediune 19, 1991.Accepted for publication November 19, 1991.

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from

PLASMA CAROTENOID RESPONSE IN MEN 1121

TABLE 1Description of treatment groups by amount and type of carotenoid supplementation*

Carotenoid content

TotalTreatment group /3-carotene a-carotene Lutein Lycopene carotenoids

mg

30 mg CarotenoidsPurified fl-carotene (n = 5; capsule) 30 ND ND ND 30Carrots (n = 5; 272 g) 29 9 Trace ND 38

12 mg CarotenoidsPurified fl-carotene (n = 5; capsule) 12 ND ND ND 12Tomato juice (n 5; 180 g) Trace ND ND 12 12

6 mg CarotenoidsBroccoli (n = 5; 300 g) 3 ND 3 ND 6

Placebo (n = 5; capsule) ND ND ND ND ND

* ND, not detectable; trace, < 1 mg.

preparation (cooked) in which they were served to study partic-ipants. Multiple analyses of these samples of broccoli, carrots,and tomato juice were performed by using the methods ofKhachik and Beecher (8) and Khachik et al (9). The quantitiesgiven to each group of subjects were determined on the basis ofthe chemical analyses ofthe vegetables and provided the desireddaily carotenoid amount (Table 1).

Purified /3-carotene sourceThe /3-carotene capsules contained dry gelatin beadlets of 10%

/3-carotene compound with BHT, BHA, and sodium benzoateas preservatives (Hoffmann-LaRoche, mc, Nutley, NJ). Theplacebo contained beadlets without /3-carotene. Multiple analysesconducted in our laboratories confirmed the dose and compo-sition of the synthetic capsules.

Controlled diets

Each participant consumed daily one ofthe six treatments atmeals (split between lunch and dinner) together with a strictlycontrolled diet for 6 wk (42 d), followed by a 4-wk posttreatmentperiod with a self-selected diet. The treatment foods were cookedaccording to instructions on the package from the supplier (eg,precooked, flash-frozen foods were boiled in serving bags). Thecontrolled diet was formulated to have a constant and low totalcarotenoid content (0.5-1.6 mg/d for the 12 552-J intakeamount), and to exclude carotenoid-containing foods used astreatments (eg, carrots, broccoli, and tomatoes). All meals wereprepared under the supervision of a dietitian at the HumanStudies Facility of the Beltsville Human Nutrition ResearchCenter (BHNRC) at the US Department of Agriculture inBeltsville, MD. Breakfast, lunch, and supper were consumed inthe dining facility on weekdays. All meals on weekends andholidays were prepared for carryout.

Seven days ofmenus were formulated from commonly avail-able foods and were provided at 836-J increments between10 878 and 16 736 J/d. Participants were started at estimatedcaloric intakes on the basis ofstature and weight, and adjustmentswere made as needed to maintain constant weight throughoutthe study. All daily diets, regardless ofthe calories they provided,had fixed percentages of calories from fat (40%), carbohydrate

(43%), and proteins (1 7%). Alcohol was not permitted duringthe controlled-diet period but instant coffee and diet lemon-lime, carbonated soft drinks were allowed ad libitum throughout.

P/asma analysesFasting blood samples were collected 1 wk before the start of

the study, at baseline, and twice weekly during the 42.4, con-trolled-diet portion of the study for determination of plasmacarotenoids, retinol, a-tocopherol, high-density-lipoprotein(HDL) cholesterol, low-density-lipoprotein (LDL) cholesterol,and triglyceride values. Blood collection continued at 7, 14, and28 d posttreatment for all study participants.

Blood for carotenoid analyses was collected in all-plastic sy-ringes, with 4500 u heparin/L whole blood. All samples wereprotected from light and were centrifuged at 2260 X g for 20mm at 12 #{176}Cwithin 1 h after blood drawing. Plasma sampleswere stored at -70 #{176}Cand were not thawed until analysis 17-19 mo later. Individual carotenoids and their total concentrationwere analyzed by HPLC methods ofBieri et al (10) as describedby Craft et al (I I). This analytical method measures 90% of thetotal plasma carotenoids present by summing the seven mdi-vidual carotenoid peaks observed, including a-carotene, /3-car-otene, lutein/zeaxanthin, precryptoxanthin, cryptoxanthin, ly-copene, and unidentified carotenoids. Crystalline a-carotene, /3-carotene, hycopene (Sigma, St Louis), zeaxanthin, cryptoxanthin,and echinenone (Hoffmann-LaRoche, Inc, Nutley, NJ) were usedas standards.

A pooled plasma reference material was analyzed daily inconjunction with all samples analyzed. Coefficients of variationfor the seven carotenoid peaks that were summed and reportedas total carotenoids varied from 2.8% (zeaxanthin/hutein) to10.0% (precryptoxanthin).

Although the seven carotenoid fractions obtained were usedto calculate total carotenoids, only the four major peaks (a-carotene, /3-carotene, hycopene, and lutein/zeaxanthin) are re-ported in detail here. Triglycerides, cholesterol, and HDL cho-lesterol were analyzed on a Centrifichem Analyzer (BakerInstruments Corp, Allentown, PA) by using standardized pro-cedures provided by the manufacturer. LDL cholesterol was es-timated by the formula of Friedewald et al (12). Plasma retinoland a-tocopherol were determined by HPLC (13).

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from

MICOZZI ET AL

nm

CS

L

303.1 130.578.5 53.7

409.6 131.1892.3 336.1

2139.0 564.7

2.31 0.4622.74 5.49

4.09 0.771.26 0.222.53 0.750.66 0.2223.4 1.6

-7 0 7 14 21 28 35 42 49 56 63 70

SSD Control Diet Self Selected Diet (SSD)

The 30 subjects entered into the study weighed between 64.1

n

m

ISSIL

1.200

1,000

800

600

400

n

m

IeS

L

SSD Control DIet

-7 0 7 14 21 28 35 42 49 56 63 70

DaysSelf Selected DIet ($80)

1122

TABLE 2Mean baseline concentrations of plasma carotenoids, and biochemicalindices for study participants

Carotenoids (mmoh/L)a-carotenea-carotene

Lutein/zeaxanthinLycopeneTotal carotenoids

Biochemical indicesRetinol (mol/L)Tocopherol (jzmoh/L)Cholesterol (mmoh/L)HDL cholesterol (mmol/L)LDL cholesterol (mmoh/L)Triglycerides (mmol/L)Body mass indext

* I SD; n = 30.t Weight (in kg)/stature2 (in m2).

Statistical analyses

Summary statistics were calculated for the carotenoids (re-sponse variables) for each of the six groups. The distributionsof response variables were compared by using the I test for twoindependent groups, analysis of variance (ANOVA) for morethan two independent groups, and nonparametric counterparts(the Wilcoxon two-sample rank-sum test and the Kruskal-Walhistest) (14). Comparisons of means were based on Duncans mul-tiple-range test as calculated with SAS (1 5). Paired t tests andthe signed-rank tests were used to compare for significant changesabove baseline.

Results

and 88.2 kg and had baseline plasma total carotenoid concen-trations that averaged 2139 565 nmol/L. Baseline values ofcarotenoids and other biochemical indices at the start of thecontrolled-diet are shown in Table 2.

Results ofplasma analyses for the four major carotenoids andfor the total carotenoid concentration in each of the six groups

Days

FIG 2. Mean plasma a-carotene response to foods and supplements.0, 30-mg fl-carotene capsules; X, 12-mg fl-carotene capsules; +, carrots;, broccoli; 0, tomato juice; and , placebo.

are shown in Figures 1-5. There were no dramatic changes incarotenoid concentrations during the week of self-selected dietsbefore the start of the treatment period.

Table 3 shows the average maximum change in carotenoidconcentrations over the treatment period. Beta-carotene con-centrations increased significantly from baseline values with in-takes of 30- and 12-mg capsules and with carrots. No significantincreases in /3-carotene concentrations were seen in the othertreatment groups. The average maximum change was greatestin the 30-mg-capsule group, followed by the response ofthe 12-mg group and the carrot group. These three groups differed sig-nificantly from each other and all had a significantly higher av-erage maximum change than did the placebo group.

Alpha-carotene concentrations increased to the greatest extentwith carrots and the average maximum change was significantlydifferent from all the other treatment groups. Although a-car-otene in the /3-carotene-supplement groups appeared to increase,only the 30-mg-capsule group had an average maximum changesignificantly different from that of the placebo group.

The average maximum change in lutein concentration in thecarrot and tomato-juice groups did not differ from that of theplacebo group. In all three groups the lutein concentration in-creased from baseline. However, the average maximum changein lutein concentration in the supplemented groups decreased

FIG 1 . Mean plasma j9-carotene response to foods and supplements.0, 30-mg p9-carotene capsules; X, 12-mg /3-carotene capsules; +, carrots;ts, broccoli; 0, tomato juice; and , placebo.

200

0-7 0 7 14 21 28 35 42 49 56 63 70

SSD Control Diet Self Selected Diet (SSD)Days

FIG 3. Mean plasma lutein response to foods and supplements. 0,30-mg /3-carotene capsules; X, 12-mg /3-carotene capsules; +, carrots; ,broccoli; D, tomato juice; and , placebo.

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from

1,600

1,400

n

m

IeSIL

600

400

200

-7 0 7 14 21 28 35 42 49 56 63 70

580 Control Diet Self Selected Diet (880)Days

HG 4. Mean plasma lycopene response to foods and supplements.0, 30-mg /3-carotene capsules; X, 12-mg /3-carotene capsules; +, carrots;f2, broccoli; El, tomato juice; and , placebo.

10.000

8.000

6,000

4,000

n

m

ISSIL

-7 0 7 14 21 28 35 42 49 66 63 70

850 Control DIet Self Selected DIet (880)Days

FIG 5. Mean plasma total carotenoids response to foods and supple-ments. 0, 30-mg /3-carotene capsules; X, 12-mg /3-carotene capsules; +,carrots; z, broccoli; 0, tomato juice; and , placebo.

PLASMA CAROTENOID RESPONSE IN MEN 1123

from baseline and was significantly lower than that in the placebogroup; the broccoli group had a significantly greater averagemaximum change than all groups.

During the treatment phase, plasma lycopene concentrationdeclined consistently from baseline in all groups except the to-mato-juice group. The tomato-juice group had a slightly positiveaverage maximum change that was significantly different fromall other groups.

The average maximum change in total carotenoid responsewas largely determined by increases in /3-carotene concentrationsin the /3-carotene-supplemented and carrot groups. The averagemaximum change in total carotenoid concentration in the broc-cohi and tomato-juice groups did not differ from that in the pla-cebo group.

No treatment group reached plasma steady state for /3-caroteneor total carotenoids insofar as plasma concentrations were stillrising at 42 d and declined immediately after cessation of treat-ment, simulating a peak on the final day of treatment. The ca-rotenoids that increased substantially during the treatment perioddecreased slowly during the 28 d of posttreatment observation.Some carotenoids, especially /3-carotene in the capsule groupand a-carotene in the carrot group, did not return to baselinevalues. Lycopene immediately increased after the treatment pe-nod in all groups except the tomato-juice group.

Figure 6 shows the mean plasma concentrations of carotenoidsin the placebo group. Lycopene fell precipitously during thecontrol-diet period and responded markedly after change to self-selected diets. In contrast, lutein increased whereas a and /3-carotene remained relatively constant.

Discussion

When the study population was consuming self-selected diets,both plasma hycopene and lutein/zeaxanthin made up a greaterpercentage of the total carotenoids (43% and 19%, respectively)than did /3-carotene (14%). Any suggested cancer-preventionagent in foods would, after consumption, probably be requiredto appear in the blood to a significant extent to have an effecton cancer sites other than those ofthe gastrointestinal tract (7).

We designed this study with a defined diet to control for dietaryinfluences on absorption and utilization. Dimitrov et al (16)showed that dietary fat (63-69 g from breakfast and lunch plus

an unreported quantity for the self-selected evening meal) affectsthe plasma response to /3-carotene supplements. Our diet had40% of the jouhes from fat, 17% from protein, and 43% fromcarbohydrates; it was moderately low in total carotenoids (0.5-1 .6 mg/d at the energy intake of 12 552-J). Furthermore, thetreatment doses were strictly monitored throughout the studyto ensure compliance. The plasma samples were stored for 17-19 mo at -70 #{176}Cwithout thawing before analysis. Under iden-tical conditions in a controlled storage study, we observed nosignificant deterioration in carotenoid concentrations for 28mo (I I).

Plasma /3-carotene concentrations increased dramatically inresponse to purified /3-carotene. Our data confirm earlier studiesdemonstrating a substantial increase in serum or plasma /3-car-otene in response to purified /3-carotene at concentrations as howas 12 mg for 42 d (16, 17). Previously we reported carotenoder-mia in subjects ingesting the 30-mg capsules during this study(18). Constantino (17) noted a dramatic increase in fl-caroteneconcentration after 2 mo of 15 mg /3-carotene/d followed by aplateau in concentration after 4 mo. Plasma /3-carotene concen-tration in our supplemented groups (30 and 12 mg /3-carotene/d) also showed a sharp initial increase and may have stabilizedif the supplementation had continued for a longer time.

Although the large serving (272 g) ofcarrots provided 29 mg/3-carotene/d, the average maximum change in plasma concen-tration was only 18% of that in the group that received 30-mgcapsules of /3-carotene. Previous investigators noted that carot-enoids in foods are much less available than are purified sourcesof /3-carotene (19, 20).

It is interesting to observe that a-carotene concentrations in-creased in the purified /3-carotene groups whereas a-caroteneconcentrations did not increase in the placebo group. This in-crease could not be due to the trace amount ofa-carotene foundin the purified /3-carotene by chemical analyses. We speculatethat this may be an analytical artifact because a- and /3-caroteneare not completely resolved to baseline concentrations whenplasma contains high concentrations of/3-carotene and as a resultvalues for a-carotene may be increased.

Carrots provided a good source of both a- and /3-carotene.Carotenes from cooked carrots are absorbed intestinally moreefficiently than are those from raw carrots because the plant cellsare disrupted during preparation and because the mechanicalhomogenization ofcarrots allows greater efficiency of absorption

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from

IDays

TABLE 3

1124 MICOZZI ET AL

S SD. Means in the same column with different superscripts are significantly different, P < 0.05.

(2 1). Alpha-carotene concentrations increased (in nmol/L permg carotenoid) twice as much as those of /3-carotene when themaximum change from the dose given was compared. Theseresponse differences may result from the more efficient conver-sion of /3-carotene than of a-carotene to retinol (22). Alterna-tively, the absorption of the a-carotene may have been greaterthan the absorption of /3-carotene from the same food source.Kim et al (23) also observed a greater serum-dose response fora-carotene than for /3-carotene in subjects receiving carrot juice.

Lutein concentrations in the broccoli group increased signif-icantly over time, reflecting the 3 mg lutein/d derived frombroccoli; the maximum change of679 nmol/L was significantlygreater than the increase in the placebo group. In contrast, themaximum change in plasma /3-carotene from broccoli did notdiffer significantly from that in the placebo group.

Plasma lutein concentrations increased when the placebo andthe controlled diets were fed (Fig 6). However, when large dailydoses ofpunfied /3-carotene were given (12 or 30 mg), the plasmalutein response was significantly lower than the response to pla-cebo. Large doses of purified /3-carotene may impair intestinalabsorption oflutein. This suggestion of carotenoid interactionsshould be considered when the longer-term effects of /3-carotenesupplementation are evaluated.

Baseline lycopene concentrations in this predominantly col-lege-aged study population were more than double those reportedin a population of middle-aged men (24). The rapid decreaseduring the treatment period and increase afterwards for all groupsexcept the tomato-juice group indicates that the self-selecteddiet ofour subjects contained substantially more lycopene thandid the controlled diet. Because tomatoes are the major con-tnbutor of dietary lycopene compared with other foods in theUS diet (7), these observations suggest high intakes of tomatoesand/or tomato-containing foods. We are currently examiningthe food histories of our subjects to determine the types of foodeaten and carotenoid intakes. Based on the lycopene concentra-tion at day 0, the calculated half-life for all groups except thetomato-juice group was 1 1-14 d. When the natural logs of thelycopene concentration ofthese groups were ploued against time,only the first 14 d ofdechine were linear, indicating the existenceof more than one body pool.

Although the changes in total carotenoid concentrations re-flected the dramatic increases in /3-carotene concentrations, moresubtle changes in other carotenoids were not readily apparent.Significant changes in carotenoid concentrations in the broccoli-and tomato-treatment groups did not result in total carotenoid

FIG 6. Mean response of plasma carotenoids to placebo capsule. #{149},/3-carotene; 0, a-carotene; , lutein; and #{149},lycopene.

concentrations that were significantly different than those in theplacebo group. These results emphasize the need for sensitivemethods such as HPLC to detect alterations in individual ca-rotenoid.

Although the controlled diet was moderately low in calculatedcarotenoid content, the placebo group did not show any dramaticchanges over the treatment period except for lycopene. Thissuggests that plasma carotenoid concentrations other than ly-copene remain fairly constant and are slow to change with lowdietary amounts. The gradual return toward baseline after sup-plementation for /3-carotene in the capsule and carrot groupsand in a-carotene in the carrot group indicates a slow turnovertime for these carotenoids. For the period after supplementationwe calculated a half-life of 7-14 d for /3-carotene in the capsuleand carrot groups and in a-carotene in the carrot group. Whenthe natural hogs of the carotenoid concentrations were plottedagainst time, each of the four graphs was linear during the first14 d followed by a break or change in slope. This nonlineardecline indicates the existence ofmore than one body pool, sim-ilar to our observations for lycopene.

There appeared to be less variability in the plasma responseto the continual supplementation of dietary carotenoids in thisstudy population than there was in a different study populationin our controlled study of a single dose of dietary carotenoids(25). With continual carotenoid supplementation, all participantsin a treatment group showed plasma carotenoid responses in acontinuum with other members of that group.

Despite dramatic increases in /3-carotene in some groups, therewere no significant changes in retinol concentration over timein any ofthe treatment groups. These data confirm observations

Average maximum change in plasma concentrations of carotenoids, by type of supplementation5

TotalTreatment group /3-carotene a-carotene Lutein Lycopene carotenoids

nmo//L

3Omg/3-carotene(n=5) 7901 138l 28875 -l5248 7344 178112 mg /3-carotene(n = 5) 3589 707k 98 46 -22 105.b -756 217 2977 647kCarrots(n = 5) 1438 762C 971 377C 381 257 747 247 2366 l542IBroccoli(n=5) 193 l32l -l33l 679 151 -778215 302465cTomatojuice(n=5) h793d 90 107b.e 34306b 237567cPlacebo(n=5) 21111 233199ce -612234

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from

PLASMA CAROTENOID RESPONSE IN MEN 1125

by others that circulating retinol does not appear to be affectedby long-term doses ofpurified /3-carotene in subjects with normalbaseline concentrations (26).

Overall, the plasma response to purified /3-carotene was greaterthan it was to similar quantities ofcarotenoids from food. Thesemicronutrient supplements appear to be more readily absorbedthan are the same micronutrients in foods, in terms of producinga plasma response. We did not measure metabolic indices otherthan plasma response, however. Given the marginal plasma re-sponses observed to some foods, it may be difficult to markedlyraise the blood concentrations of some carotenoids when com-mon foods are the sole source. Participants in the present studyattested that it would be difficult to physically consume dailyamounts ofcarrots (272 g) or broccoli (300 g) much greater thanthe amounts provided in the study diets. Levels of consumptionof tomato juice (180 g/d), however, could have been markedlyincreased without gastric discomfort. Consumption of tomatojuice at amounts up to 2 L/d for several years have been reported,resulting in lycopenemia (27). This level ofconsumption is morethan 10-fold that in the present study. In addition, among thegeneral population, some individuals may have low carotenoidconcentrations irrespective of diet. Differences in absorption,transport, and efficiency in retinol conversion may affect plasmacarotenoid concentration.

To achieve major increases in specific plasma carotenoid con-centrations, it may be more effective to ingest purified carot-enoids than to attempt major alterations in the human diet.However, foods high in carotenoids can contribute smaller butsignificant increases in certain carotenoids and also provide manyother biologically active constituents that are not present in pu-rifled micronutrient supplements. These constituents also maybe important for decreasing the risk ofcancer and/or contributingto optimal health (7).

Better understanding of carotenoid metabolism and the bio-availability ofcarotenoids from foods prepared in different wayswill be important to formulating meaningful recommendationsregarding dietary carotenoid consumption. a

We thank Alice Rose, Neal Craft, Karen Seidel, and the staff of theVitamin and Mineral Nutrition Laboratory, USDA; Priscilla Steele andthe staff of the Human Studies Facility, USDA; William Campbell ofthe Cancer Prevention Studies Branch, NCI; and David Annett, RobertBanks, and Robert Smucker of Information Management Services,Inc, for their technical assistance in this study. Hemmige Bhagavan,Hoffmann-LaRoche, Inc, Nutley, NJ, provided samples of purifiedcarotenoid standards as well as helpful consultative advice.

References

1. Peto R, Doll RK, Buckley JD, Spore MB. Can dietary beta-carotenematerially reduce human cancer rates? Nature 198 1;290:201-8.

2. Hennekens CH, Maryrent SL, Willett WC. Vitamin A, carotenoidsand retinoids. Cancer l986;58: 1837-41.

3. Micozzi MS. Foods, micronutrients and reduction ofhuman cancerrates. In: Moon TE, Micozzi MS, eds. Nutrition and cancer preven-tion: investigating the role of micronutrients. New York: MarcelDekker, 1988:213-41.

4. Ziegler RG. A review of epidemiologic evidence that carotenoidsreduce the risk ofcancer. J Nutr l989;l 19:1 16-22.

5. Albert-Puleo M. Physiologic effects ofcabbage with reference to itspotential as a dietary cancer inhibitor and its use in ancient medicine.J Ethnopharmacol l983;9:261-72.

6. Riddle JM. Ancient and medieval chemotherapy for cancer. Isis1985;76:3 19-30.

7. Micozzi MS, Beecher GR, Taylor PR, Khachik F. Carotenoid anal-yses of selected raw and cooked foods associated with a lower riskfor cancer. JNCI l990;82:282-5.

8. Khachik F, Beecher GR. Decapreno-beta-carotene as an internalstandard for the quantification of the hydrocarbon carotenoids byhigh-performance liquid chromatography. J Chromatogr 1985;346:237-46.

9. Khachik F, Beecher GR, Whittaker NF. Separation, identificationand quantification of the major carotenoids and chlorophyll con-stituents in extracts of several green vegetables by liquid chroma-tography. J Agric Food Chem l986;34:603-16.

10. Bieri JG, Brown ED, Smith JC. Determination ofindividual carot-enoids in human plasma by high performance liquid chromatog-raphy. J Liq Chromatogr l985;8:473-84.

1 1 . Craft NE, Brown ED, Smith JC. Effects of storage and handlingconditions on concentrations ofindividual carotenoids, retinol andtocopherol in plasma. Gin Chem 1988;34:44-8.

12. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the con-centration oflow density hipoprotein cholesterol in plasma withoutuse of the preparative centrifuge. Gin Chem h972;18:499-502.

13. Bieri JG, Tolhiver Ti, Catignani GL. Simultaneous determinationofalpha-tocopherol and retinol in plasma or red cells by high pressureliquid chromatography. Am i Clin Nutr 1979;32:2l43-6.

14. Snedecor GW, Cochran WG. Statistical methods. 6th ed. Ames, IA:Iowa State University Press, 1967.

15. SAS Institute, Inc. SAS Users Guide: Statistics, Version 5 Edition.Cary, NC: SAS Institute, Inc, 1985.

16. Dimitrov NV, MeyerC, UlIrey DE, et al. Bioavailability of /3-carotenein humans. Am i Chin Nutr 1988;48:298-304.

17. Constantino iP, Kuller LH, Begg L, Redmond CK, Bates MW.Serum level changes after administration of a pharmacologic doseof beta-carotene. Am i Chin Nutr 1988;48: 1277-83.

18. Micozzi MS, Brown ED, Taylor PR, Wolfe MA. Carotenodermiain men with elevated carotenoid intake from foods and fl-carotenesupplements. Am i Clin Nuts l988;48:l061-4.

19. Rao CN, Rao BSN. Absorption ofdietary carotenes in human sub-jects. Am i Chin Nutr l970;23: 105-9.

20. Anonymous. Requirements ofvitamin A, thiamine, riboflavin andniacin. FAO Nutr Meet Rep 5cr (h967;4h :74-6. (WHO technicalreport series 362.)

21. Van Zeben W, Hendriks TF. The absorption ofcarotene from cookedcarrots. Int Z Vitaminforsch 1948;l9:265-6.

22. Bieri iG, McKenna MC. Expressing dietary values for fat-solublevitamins: changes in concepts and terminology. Am i Chin Nutr198 1;34:289-95.

23. Kim H, Simpson KL, Gerber LE. Serum carotenoids and retinol ofhuman subjects consuming carrotjuice. Nutr Res 1988;8:1h19-27.

24. Stacewicz-Sapuntzakis M, Bowen PE, Kikendall iW, Burgess M.Simultaneous determination of serum retinol and various carot-enoids: their distribution in middle-aged men and women. i Mi-cronutr Anal l987;3:27-45.

25. Brown ED, Micozzi MS, Craft NE, et al. Plasma carotenoids innormal men after single ingestion of vegetables or purified beta-carotene. Am I Chin Nutr 1989;49: 1258-65.

26. Willett WC, Stampfer Mi, Underwood BA, Taylor JO. Vitamin A,E and carotene: effect of supplementation on their plasma levels.Am i Chin Nutr 1983;38:559-66.

27. Reich P, Schwachman N, Craig iM. Lycopenemia: a variant of car-otenemia. N EngI i Med 1960;262:263-9.

by guest on March 6, 2014

ajcn.nutrition.orgD

ownloaded from