Folate, Vitamin C, And Cervical Intraepithelial Neoplasia

Vol.

1 , I 19- 124. /anuarv/1ehruarv

Folate,

Juliet VanEenwyk,2

Cancer

1992

Vitamin

Faith G. Davis,

C, and Cervical

and Neville

Colman

Department of Epidemiology and Biostatistics, School of Public Health, University of Illinois at Chicago, Chicago, Illinois 6061 2 [1. V., F. G. D.]; and VA Medical Center, Bronx, New York 10468, and Center for Clinical Laboratories, Mount Sinai Medical Center, New York, New York 10029 [N. C.]

Abstract A case-control study was designed to assess the relationship between cervical intraepithelial neoplasia (CIN) and folate in serum, red blood cells, and diet. The association between CIN and dietary vitamin C was also investigated. Cases were selected from women with biopsy-confirmed CIN. Controls were age-, race-, and clinic-matched women with normal cervical (Pap) smears. Study participants completed selfadministered food frequency (n = 100 matched pairs) and health (n = 102 matched pairs) questionnaires. Fasting venous blood samples were collected for serum (n = 98 matched pairs) and red cell (n 68 matched pairs) folate assays. Conditional logistic regression models were used to estimate crude odds ratios and odds ratios adjusted for smoking, income, number of sexual partners, frequency of cervical smear, use of spermicidal contraceptive agents, history of genital warts, and Quetelet index. Dietary intake variables were adjusted for total energy intake prior to logistic regression. A protective effect of red cell folate was evident with adjusted odds ratios (95% confidence intervals) of 0.1 (0.0-0.4), 0.6 (0.2-2.0), and 0.5 (0.21.9) for those in quartiles 4 (highest), 3, and 2 compared to quartile 1 (lowest). Supporting evidence for the protective effect of folate was provided by inverse associations between CIN and folate in both serum and diet. An inverse association was also found between CIN and dietary vitamin C with adjusted odds ratios (95% confidence intervals) of 0.2 (0.0-0.7), 0.6 (0.2-1.6), and 0.6 (0.2-1.8) for those in quartiles 4, 3, and 2, respectively, compared to quartile 1. These findings support dietary recommendations, such as those of the American Cancer Society, the National Cancer Institute, and the U.S. Dietary Guidelines, which allow for adequate intake of folate and vitamin C, both of which are found in good quantity in fruits and vegetables. Increased consumption of legumes and whole grains is also in accord with current dietary

Re eived 4/t 6/91. 1 Funded in part by a grant froni the American Canc er Society, Illinois Division, and I)y the State Cancer Plan of the Illinois Cancer Council. 2 To whom requests for reprints should be addressed, at Illinois Department of Public Health, Division of Health Statistics and Policy Development, 1 01) West RailoIph, Suite 6-600, Chicago, IL 60601.

Epidemiology,

lntraepithelial

Biomarkers

& Prevention

1 19

Neoplasia1

recommendations, and both of these types of foods are good sources of folates. Introduction Folic acid, a B-group vitamin, is necessary for normal cell replication, and cells grown in folate-deficient media manifest chromosomal abnormalities which correspond to those found in many types oftumor cells (1). In relation to cervical cancer, it has been noted that folate deficiency

can lead to cervical cellular changes which resemble neoplastic change (2) and that preneoplastic cervical cellular changes among users of oral contraceptives megness with

folate

supplementation

(3). Three

studies of folate consumption have found little evidence folate

intake and disease. This investigation assesses

folic

acid

chosen

and

to attenuate from rather

ensue pothesized and with

CIN.3

The

the

relationship

premalignant

physiological than foreshadow

that higher

case-control

and cervical cancer (4-6) of an association between

levels

between

condition

changes disease.

of serum

was

which may It was hy-

and red cell folate

higher dietary intake of folate would be associated a reduction in disease. Because folates and vitamin

C are found

in many

of the

same

foods

and

vitamin C protects folates from oxidative tary intake of vitamin C was also assessed.

Materials Details

because

cleavage,

die-

and Methods of case and control

measurement procedures, variables, and statistical

selection,

serum

measurement analysis have

and dietary

of confounding been previously

documented (7, 8). These topics are reviewed below. Case and Control Selection. Participants were recruited from

clinics

Illinois

at Cook

Hospital

County

between

(n = 102) were selected with biopsy-confirmed

matched tended smears

controls the same showed

April

Hospital

and

from

women

aged

CIN

I, II, or

Ill.

were selected clinics as the

no abnormality

University

of

1987 and June 1 989. Cases

from cases

18 to 49

Age-

and

years

race-

women who atand whose Pap

of a severity

greaten

than

or equal to benign atypia. Women who had been pregnant or lactating within the past year were excluded from the study because of the potential for folate depletion under these circumstances. In this population, pregnant and postpartum women may also be at increased risk for a diagnosis of CIN, since pregnancy brings women into the clinic where cervical smears are obtained as part of prenatal care. Women with epilepsy on sickle cell anemia were also excluded, since these conditions are associated with low blood folate and with bringing women into the

‘ (

The abbreviations rude odds ratio;

used are: CIN, cervical intraepithelial OR,,, adjuste(l odds ratios; CI, ci)ntidenc(’

neoplasia; interval.

OR,.

120

Folate,

Vitamin

medical

C, and Cervical

system

smears.

where they with diabetes

Women

requirenient All

the

for eligible

ci 1 0-h

Neoplasia

likely to have cervical were exduded due to the

are

f,ist.

wonien

of

study.

lnlraepithelial

were

requested

to

participate

in

eligible cases, 102 were enrolled, yielding a panticipation rate of approximately 61%. To enroll an equal number of controls, 195 eligible women were approached, giving a participation rate of 52%.

Measurement

166

of Exposure.

The food frequency portion History Questionnaire of the Division of Cancer Prevention

of the Health Habits and National Cancer Institute,

and Control,

version

2.

1 , was

used

to assess

dietary

from

this procedure

includes

vitamin

C from

both

food and vitamin supplements. There is no provision for including supplemental folates in the dietary folate measure. Adjustment for total energy intake using the regression procedure of Willett and Stampfer (10) was used to control for overand underreporting of dietary intake. Two participants failed to adequately complete the food frequency questionnaire, resulting in 100 matched pairs

available

for analysis.

Fasting

radioassay

venous

blood

of serum

and

lysates were prepared (11). Serum was also samples were stored laboratory on dry ice

samples

red cell

were

folate.

collected

Red cell

on site by the method of Gutcho aliquoted on site, and all blood at -70#{176}C until shipment to the every 6 to 10 weeks. Assays for

of variation was limits of sensitivity

levels were associated variation. All laboratory

with the personnel

questionnaire and pregnin

ual

Inforniation

behavior.

RIrti( pants (ompleted a whi h asked aI)out backy history, smoking, and sexmroni

this

questionnaire

was

used to assess independent contributors to risk of CIN in this sample and to control for confounders of the diseaseexposure relationship. Confounders were defined as vanables which have been reported as risk factors in previous studies and variables whose inclusion led to a change of more than 20% in the adjusted odds ratio for the nutrients of interest.

Analysis. OR, and OR,, and 95% CIs were estimated using the MCSTRAT program (15), which performs an iterative conditional maximum-likelihood fit of a logistic regression model. Quartiles for the hematologcal measures

3.8-8.2% for conof the assay. Low

highest coefficients of were unaware of case

or control status of the blood samples. Failure to withdraw blood from four women resulted in serum measures for 98 matched pains. Inadequate on-

site preparation of the red cell hemolysates resulted in unreliable data from the first 28 cases and 14 controls, leading to the exclusion of 28 matched pairs from the final red cell folate analyses. Failure to collect a lavendertop tube from one woman and missing red cell data from one woman resulted in the exclusion of an additional two pairs, leaving 68 matched observations for the red cell folate analyses. Because more cases were enrolled at the beginning of the study and more controls enrolled toward the end of the study period, there was a disparity in the allocation

and calorie-adjusted

nutrient

intake

were

defined from the distribution of the controls. Those in the lowest quartile (quartile 1) served as the comparison group. Adjusted models included independent contributors to risk in this sample, as well as potential confoundens of the disease-exposure relationship. Tests of trend were achieved by entering quartiles of a given nutrient into the logistic model as different values of a single ordinal variable. Pearson product-moment correlation coefficients for the correlation between the natural log of the hemato-

logical

measures

and calorie-adjusted

intake

measures

were

for

hemo-

serum and red cell folate were conducted by modiuication of the methods of Waxman et a!. (12) and Longo and Herbert (13), respectively, using Becton Dickinson Simultrac kits as reagents (14). For the folate assays, the intraassay coefficient of variation was 1 .4-4.6%, and the intenassay coefficient trol samples at the

of Confounders.

self-administered ground, health

Statistical

intake

of folate and vitamin C (9). Participants were asked to complete this questionnaire prior to the clinic visit. The conversion of foods on the food frequency questionnaire to nutrients was accomplished via the microcomputer software version 2.2, August 1989, provided by the National Cancer Institute, Division of Cancer Prevention and Control (9). The measure of dietary vitamin C ob-

tamed

Measurement

Confounding

generated

due

nutrient using

to intenassay

and food

SAS procedures.

variability

was

as-

sessed models.

by including a dichotomous variable in the logistic This variable was (Treated by calculating the mean value for the quality control samples and charactenizing assay groups according to whethen their quality control samples were above on below the mean.

Results The

distribution

and nondietary been presented

of cases

and

risk factors (7, 8). Table

controls on demographic associated with CIN have 1 shows the OR,,s and 95%

CIs for the nonnutnient variables included in the multivanial)le conditional logistic regression models. Increased OR,,s were associated with current smoking status, more than t’te year between cervical smears, any use of contraceptive spermicidal foams or gels, and a self-reported history of genital warts. An inverse association was observed between OddS of disease and monthly income bracket in $400 increments to $2000. Quartile of Que-

telet

index

(kg/m2)

and

number

of sexual

partners

were

100% of red cell samples included in the logistic models. Laboratory personnel were unaware of the inclusion of

not independent contributors to risk after adjustment for the other variables. However, these variables were metamed in the final model, because they were considered to be potential confoundens of the disease-exposure relationship. Use of oral contraceptives and parity have been reported to relate to both folate status and risk for CIN. However, since these factors were not independent contributors to risk in this sample and their inclusion in the logistic models did not alter the adjusted estimates, they were not included in the final models. Excessive alcohol consumption is associated with lowered blood folates. Controlling for this variable (lid not alter the findings and it is not included in the adjusted models. Table 2 shows the quartiles of serum and red (Tell folate. The number of cases and controls in each quartile

these

and the percentage

of case groups.

and control To enable

blood samples to the laboratory control for confounding due

tween-nun variability, two pooled blood were included These 1 1 batches accounted

samples.

assay to be-

quality control samples of in 1 1 of the 16 shipments. for 88% of the serum and

with

deficiencies

are also presented.

Cancer

Table

1

Adjusted included .

Nonnutrient Current (yes

odds ratios and 95% in all adjusted logistic .

variable

95%

.

Odds

smoker versus

CIs for nonnutrient regression models ratio

confidence interva 1.2-5.5

0.5

0.3-0.8

no)

Income bracket (ordinal, monthly $400 increments

dietary

Frequency of cervical smear (less often than annual versus annual)

3.3

1 .3-8.2

Use of contraceptive dal agents

2.8

1.3-6.3

3.4

1.1 -10.8

0.7

0.5-1.0

1.0

0.6-1.7

(ever

used

Self-reported

history

warts (yes versus Quetelet (ordinal

never

versus

Numberofsexual (log,-transformed) Odds shown.

For

ratio

partners

for

each

variable

comparison,

Second

the

National

in the

deficiency

and

all

of

of the

for folate

Nutrition

from

are also given

Evaluation

the

Survey (16).

The

poorer folate of the Second

odds

of

CIN

is observed

for

those

in

cell

folate

relative

to those

in the

the

highest

lowest

Table

2

significant

decrease

in

strongest

correlation

is between

serum

and

who paid influenced

for their own medical exthe ORs, since an under-

representation in that group could result in an overrepresentation in both higher (privately insured) and lower (public aid) socioeconomic groups. Adjustment for the nonnutrient factors in the logistic models should mitigate bias resulting from nonrandom distribution of cases and controls on socioeconomic factors resulting from differential response rates. Cases and controls were comparable on the matching variables, as well as on educational level, employment

quartile of serum folate relative to those in the lowest quartile. However, while the ORa is smaller than the ORE, the statistical significance of the decreased OR is attenuated in the adjusted model. The pattern of decreasing ORs with increasing quartile of serum folate is also attenuated after adjustment for nonnutrient factors. The decreased odds of disease for those in the highest quartile of red

the

of control women penses may have

Survey.

The OR,s, ORa5, and 95% CIs by quartile of serum and red cell folate are presented in Table 3. A decrease in

a statistically

Issues of Bias. To assess possible bias due to differential participation rates of cases and controls, information on age, ethnic origin, zip code, and type of payment for medical services was collected for all women asked to participate in the study. Response rates were lower (P < 0.05) for controls than for cases among women age 18 to 24 years and among women who paid their own medical expenses. Since cases and controls were matched on age, the effect of this discordance cannot be estimated. It is not known in what direction the deficit

variables

Evaluation

appear to manifest from the findings

C,

Discussion

other

women

and Nutrition

ranges

Health

for

percentage

Health

women in this sample status than anticipated National

is adjusted

status,

marital

traceptive

quartile

Qua rtiles of serum

status,

history

of hospitalizations,

use, and age at first intercourse.

oral

After

con-

control-

and red cell folate

Quartile

Deficienc

y levela

Nutrient

Serum folate Controls’ Cases#{176} Red

level

)ng/ml)

1

2

3

1.3-3.4 26 36

3.5-4.4 22 23

4.5-6.3 25 25

6.4-21.2 25 14

57-126

127-149 17 18

150-190 17 18

191-325 17 6

cell folate level )ng/ml)

Controls’ Cases’ a b C d

Deficiency

levels

17 26 are

those

The percentages of women Women aged 20-44 years. Estimated for women aged

e The number of cases I Estimated for women

defined from 20-64

and controls aged 20-64

by the the

years

4

Frank

Borderline

<3.0 (15%io) 14.3% 23.5%

<5.0(41%”) 61.2% 67.4%

<140

(13%’) 41.2% 51.5%

<160)20%’) 60.3%

73.5%

laboratory.

Second (Ref.

National

Health

and

Nutrition

Evaluation

Survey

who

were

below

deficiency

16, p. 37).

in each quartile and the years (Ref. 16, p. 38).

121

red cell folate. Intakes of dietary vitamin C and dietary folate are highly correlated with each other, and both of these measures seem to be derived primarily from the intake of total fruit and citrus fruit. In contrast, the hematological folate measures are modestly correlated with vegetable intake.

no)

index (kg/m2) by quartile)

vitamin

assays,

used)

of genital

& Prevention

the odds ofClN is evident in both the crude and adjusted models for those in the highest quartile of intake relative to those in the lowest quartile. Associations among the vitamin assays and dietary measures are shown in Table 5. Among the vitamin

income in to $2000)

spermici-

Biomarkers

are apparent in both the crude and adjusted models. The pattern of decreasing ORs with increasing quartile of red cell folate is also evident in both models. Table 4 presents the OR,s, ORaS, and 95% CIs for dietary intake of folate and vitamin C. Decreasing ORs with increasing quartile of dietary folate and vitamin C are evident in both the crude and adjusted analyses. For

variables

2.6

Epidemiology,

percentage

below

the

specified

deficiency

level

are

presented.

levels

(16) are

in parentheses.

122

Folate,

Vitamin

Table

I

Odds

C, and Cervical

lntraepithelial

ratios,

and tests folate

95%

CIs,

Neoplasia

ot trend

for

serum

and

red

cell T,il)le

Test

Quartile

4

Odds

ratios,

trend

i)f

95% CIs, ,uicl lists of trend Icilate .uil vit.iniin C

Nutrient 1 (low)

2

1.0

()R, 95%CI

0.9 0.4-2.0

1.0)

0.9 0.3-2.6

0.3 0.1-1.1

1.1 0.4-3.2

1.0

0.7 0)3-1.7

0.8 0.3-1.9

0.5 0.2- 1 .9

CI

Dietary OR. 95%

2

1 .0

1. 1 0.5-2.3

CI

OR, 95%CI

1.0

0.03

0.2 0.1-0.7 0.1

0.6 0.2-2.0

Dietary

high)

1 .5

P

0.5 01.2- 1.1

0.03

0.7 01.2-2.0)

0.8

0.4 0.1-1.1

0.07

vitamin

0.2 O).1 0.5

0.00”

C 1 .0)

OR, 95%

CI

OR;’ 95%

Cl

control

With

of serum

for assay

folates

substantively findings

was

attenuated.

modify ned

for

distribution

group,

the

in the OR, for those the

cell

folates.

1 .0

Because

variable

While

was

used

this procedure

signifidid

results of

the

not

or the unequal

in the assay groups, to control

crudely

for

adjusts

a

interassay

for assay

group, residual confounding may be present and the possibility of bias cannot be entirely excluded. Bias in the findings for the red cell folate may have resulted from the exclusion of the 34 pairs from the analyses. On the dietary and serum folate measures, the controls who were excluded were similar to those included. However, the distribution of excluded cases by quartile of both dietary and serum folate differed from that of cases included in the red cell folate analyses, with more than the expected number of cases in the highest

and fewer extent

that

than expected the

serum

and

in the lowest dietary

folate

cause tumors exhibit rapid cancer may be at increased Although CIN plastic process, gression from

of cancer suggested

cell multiplication

79%

of

the cases were unlikely that the

with (18).

at this stage of the disease.

Since

with CIN I or II, it ORs for those in the

s’arts, p.irtic’rs.

1 .3

-

0.2

01.6

1 .6

0.0)’

0.03

0.7

qu,irtile

of

Quetelet

index,

and

red cell folate are a result dysplastic cells. Similarly, while anorexia is often a symptom of those with cancer, it has not been documented as a feature of CIN. It is unlikely that cases exhibited either poorer folate status or lower intake of vitamin C as the result of dietary intake depression in sick patients with poor appetites. Additionally, because participants were asked to record their average frequency of consumption over the last 5 years, the potential for reported dietary intake reflecting changes subsequent to the onset of disease seems unlikely. Nonetheless, given the case-control design of this study, the possibility that the lower levels of serum folate, red cell folate, and dietary vitamin C for the cases resulted from the disease process cannot be ruled out. Levels of serum retinyl palmitate indicated high levels of compliance with the fasting requirement among both cases and controls. For the 98 pairs included in the serum folate assays, one control and two cases had levels of

retinyl

bias

due

controls

of serum

and the

by

palmitate

indicative

to

differential

regarding

Hematological

cell multiplication, those risk for folate deficiency

diagnosed elevated

0.2

sequestration

a strong,

remains that be-

of genital of sexual

quartiles

folate

quartiles.

may represent an early stage of the neothe length of time required for the proCIN I and II to cancer (19) argues against

rapid

-1 .8

(1.6

0.2

measures

reflect red cell folate, the findings for red cell folates may be overstated. Folate deficiency has been reported in patients with a diversity of malignancies (1 7). The question of whether this is a cause or a consequence unanswered. However, it has been

0.3

Adjusted br calories prior to Iogistb regression using Iine,ir regression procedures ) 1 0). The logistic moh’I inc luck’s acljustnient for current smoking status, monthly personal incoilic’, frequency of cervical smear annual versus less (ifteii), any use ut sI)ernliciclal cc)ntraceptive agents.

lower

quartile

adjustment

gradient

of cases and controls

dichotomous

statistically

in the highest This

dose

1 .6

.,

of

be limited. cant decrease

0.6

0.7

0.3

0.01

00b_04

ling for smoking and income, cases and controls were also comparable on number of sexual partners and parity. The similarity of the cases and controls on these factors and the selection of all participants from clinics serving primarily low-income individuals suggests that the sampIe was selected from a high-risk population. The ability to generalize the study findings to other populations may

seems

1)6

0)3

0.3-1.9

self-reported history natural log of number ,‘ <0.005. ( <0.05.

To the

4

0.15

Adtusted for current smoking status, monthly personal income. Irequ(’ncy of cervical sniear (annual versus less often), any use ot sperniicidal contraceptive agents, sell-reported history of genital warts, quartile of Quetelet index, and natural log of number ot sexual partners. b <0.05.

quartiles

3

folate

.,

variability.

of trend

N utrient

0.04

0.4 0.2-0.9

of

folate

1.0

OR, 95%

0.8 0.3-1.7

intake

Test

Qu artile

P

1 low)

Serum folate OR, 95%CI

Red cell OR, 95%CI

4 (high)

3

for dietary

of

fasting

among

Therefore,

cases

and

is unlikely.

and Dietary

statistically

nonfasting.

compliance

significant

Factors.

Red cell

inverse

folate

association

shows with

CIN. Additionally, the estimates of effect for both serum and dietary folate offer supporting evidence for the role of inadequate folate nutritional status in the development of CIN. Some of the difference in the findings for the folate variables may be attributable to characteristics of the measurements. Greater variability is expected in the serum compared to the red cell folate measure, because serum folate indicates short-term changes in folate balance, while red cell folate reflects changes over several months (16). For folates, there is a greater potential for misclassification from dietary intake measures compared to laboratory data because of inaccuracies in reporting and converting food to nutrients, the destruction of fo-

Cancer

Table

5

Correlationa

among

vitamin measures Folate

Red Folate Red

folate

Nutrient Folate Vitamin

intake

0.00 -0.01 0.21’

p

‘

P

<

regression 0.005.

(10)

prior

C

0.55”

0.66”

062d

065d

0.14’ 0.19’

0.12 0.12

correlations. have been Iog,-transformed. were adjusted for total

methodology

Vitamin

0.60”

energy

to correlation

preparation,

intake

using

procedure.

and differential

absorption

of folates from diverse food sources. The findings of this study are consistent with those of Butterworth et a!. (3), who noted higher levels of both serum and red cell folate in women with cervical dysplasia versus hospital employee controls. Due to the lack of statistically significant results possibly related to the

relatively

small

Butterworth’s

sample

noted an inverse cervical carcinoma

remain

retinol,

vitamin

given

the

measures

in the risk factors,

et a!. (4)

generally

high

5 and

model

folate

must

correlation

Ref.

model, as dietary

However,

of dietary

regression

(Table

adjusted as well

C, and energy. addition

a logistic

Brock

between folate intake and in situ on crude analysis. This protec-

ity of simultaneous

C into

suggestive.

association

tive effect disappeared included nonnutnient

tene,

size of 34 cases and 40 controls,

findings

vitamin

be questioned,

between

20). Of

which cano-

the validand

two

the

studies

two

on in-

vasive cervical cancer and folate intake, one (5) showed an inverse association between dietary folate and disease in an analysis which did not control for other risk factors,

and the second among heavy white women disease are not have differed present study, nonwhite

(6) revealed a protective effect for folate smokers. These two studies focused on whose socioeconomic status and risk of described. Therefore, these samples may in a significant fashion from that of the which included primarily low-income,

women

disease. Two deficiency nogenesis

from

a

population

at

high

risk

strand, proper second

which preclude the configuration of the hypothesis proposes

mechanisms

peated

of

of the many mechanisms suggested for folate causing altered DNA and subsequent carciare based on the observed misincorporation

of uracil into DNA in place of thymine, the de synthesis of which is folate dependent. One theory gests that this leads to methyl-poor regions in the

cause

excision-repair

cycles,

novo sugDNA

coiling necessary for the DNA molecule (21). The that efficient DNA repair

chromosome

breakage

which

through

aim at removing

folate

may

play

an

antitumonigenic

role

by

Evidence for an inverse association between dietary vitamin C and risk of CIN is consistent with the findings from a similar investigation (25). Of an additional three studies, one found a statistically significant inverse association between dietary vitamin C and invasive cervical cancer (5); one found no association between vitamin C intake and invasive cervical cancer (6); and one was suggestive situ with

food

123

(24).

0.05.

lates during

& Prevention

preventing preneoplastic epithelial cellular changes is suggested in both the Butterworth et a!. study (3) and the preliminary results of a chemopreventive trial with men at high risk of lung cancer (23). The findings from the latter study indicate regression of bronchial squamous metaplasia with folic acid and vitamin B12 supplementation. Folate supplementation has also been reported to be protective against the development of colon cancer and dysplasia in patients with chronic ulcerative colitis

d

0.02 0.03 0.14’ 0.19’

025d

That

y intake

Folate

Biomarkers

misincorporated uracil, but which are futile in the lowthymine environment of folate deficiency (22).

intake

0.22” 0.22”

0.14

a Pearson product-moment b Hematological measures C Dietary intake measures

dietary

Dietar

Serum

027d

Food intake Total fruit Citrus fruit Total vegetables Vegetables excluding rice and potatoes

d

cell

0.61

C

and

assays

assays cell folate

Serum

linear

assays”

Epidemiology,

me-

the

of an increase

in risk

of cervical

carcinoma

in

decreased consumption of vitamin C (4). The evidence of an inverse association between CIN and vitamin C is also consistent with the finding of lower levels of serum vitamin C in women with CIN than in age- and clinic-matched controls (26). This latter study, however, failed to control for several possible confoundens,

including

smoking.

The

suggestions

of an associa-

tion, combined with the inconsistencies of previous investigations, indicate the need for further research on the relationship ofvitamin C to cervical dysplasia. Vitamin C functions as an antioxidant and enhances cellular immunity, both of which may play a role in cancer prevention (27). Due t the relatively high correlations between the folate and vitamin C measures, it is not possible to delineate the relative importance of these nutrients as possible preventive agents in the etiology of CIN. When quartiles of dietary vitamin C and red cell folate are entered simultaneously into a logistic model with the nonnutnient variables, there continues to be a statistically significant protective effect evident for those in the highest

quartiles

of

both

measures,

and

a dose

gradient

is

evident. The decreases in odds of CIN for those in the highest quartiles of red cell folate and dietary vitamin C are also evident controlling for quartile of serum acarotene, /3-carotene, lycopene, and lutein. That dietary intake of folate is highly correlated with fruits, while the serum and red cell folate measures are more strongly correlated with vegetables, may indicate an inadequacy in the dietary folate measure. Although it was not part of the study design, serum vitamin B2 assays were conducted simultaneously with the folate assays. A preliminary review of these analyses suggests that vitamin B12 may play a role in a subset of women with cervical dysplasia. However, controlling for quartile ofvitamin B12 did not substantively alter the odds ratios associated with serum and red cell folate. The findings from this study support dietary recommendations, such as those of the American Cancer Society, the National Cancer Institute, and the U.S. Dietary Guidelines, which allow for adequate intake of folate and vitamin C, both of which are found in good quantity in fruit and vegetables. Increased consumption of legumes and whole grains is also in accord with current dietary

124

Folate,

Vitamin

C, and Cervical

recommendations, good sources

lntraepithelial

and of folates.

both

Neoplasia

of these

types

of foods

are

The authors wish Ic) acknowledge the support c)f Stanley Gall at the University of Illinois Hospital, Michael Makii at Cook County Hospital. Eileen McAleer at the Bronx V.A. Hospital, and Ray Murphy at the Illinois Department oiiiments Kviz, and

of Public Health. They have appreciated the advice and of Phyllis Bowen, lack GOIdE)erg, William Haenszel, Frederick Victoria Persky at the University of Illinois at Chicago.

References 1 . Yunis, I. I.. ,incl Siience (Washington 2. Koss, Philadelphia:

[.

C.

Soreng, A. DC), 226:

c.

E., Hatch, in cervical ontraceptives.

4. Brock. K., Berry, C., Nutrients in the diet and Cancer Inst., 80: 580-585, 5. Verreault, of diet and

and 1979.

fragile 1984.

sites

Its Histopathologic

and Bases,

Mock, P. A., Truswell, L. A., and Brinton, plasma and risk of in situ cervical caner. 1988.

Herbert,

Med.,

87:

M., and Shy, tnt. I. Cancer,

K. A case-control 43: 1050- 1054,

VanEenwyk, I. Nutrition and (Tervical lntr,sepithelial Neoplasia: Study (Ph.D. dissertation). Chicago, IL: Graduate College of Illinois, 1 990. I.. Davis, cervical

F. C., and intraepithelial

study 1989.

A Caseof the

Bowen, P. E. Dietary and serum neoplasia. Int. I. Cancer, 48: 34-

9. Block, G., Coyle, L., Smucker, R., Harlan, L., Hartman, A., and Kessler, L. Health Habits and History Questionnaire: Diet History and Other Risk Factors. Personal Computer System Packet. Bethesda, MD: National Cancer Institute, Division of Cancer Prevention and Control, 1 987-- 1 989. 10. Willett, epidemiologic

W.,

and Stampfer, analyses. Am.

1 1 . Gutcho, competitive

S. Source of error binding radioassay.

M. I. Total I. Epidemiol.,

energy intake: 124: 17-27,

in determination Clin.

Chem.,

S., Schreiber, C., and ofserum folate levels.

Herbert, Blood,

implications 1986.

of erythrocyte 24:

388-389,

for

Dilate

by

assay

for

1978.

V. Radioisotopic 38: 219-228, 1971.

V.

Raclioassav

138-151,

for

s(’rum

and

red

(

elI

1976.

and HerI’rt, folate and

V. Evaluation of siniult,ineous vitaiiin B- 1 2. Am. I. Clin. Nutr.,

Offord, K. P., S ott, W. F., md l)aoocl, S. I . The In: R. P. Hastings ed.), SUGI Supplemental Library 5 Ed., ip. 307- 328. Can,’, NC: SAS Institute, In.,

16. Sc’nti, F. R., .10(1 t’ilch, Status of the U.S. Population National Health and Nutrition MD: Life Sciences Research

ancer.

Cancer

(Phila.),

1 8. F-liii, Y. H. Hunian worth, Inc., 1983.

L. A. I. NaIl.

Control University

12. Waxman, measurement

L., and

Clin.

1 5. Naessans, I. M., MCSTRAT procedure. User’s Guide, Version 1986.

Ed. 3.

C., Brinton, L. A., Flaiiiman, R. F., Lehman, H. F., Levine, K., Norman, S. A., Rosenthal, I. F., Trumble, A. C., and Diet and the risk ot invasive cervical cancer among white United States. Am. ). Epideniiol., 112: 432-444, 1990.

8. VanEenwyk. carotenoids and 38, 1991.

D. Lab.

S. lvi. Assessnient (ii tl’ Folate Nutritional Based on Data Collected in the Sec ond Examination Survey 1976- 1 98L). Bethesda, Offices, FASEB, 1984.

1 7. Clarnon, C. H., Feld, R., Evans, W. K., Weiner, R. S., Kramer, B. S., Lininger, L. L.. Gardner, F. B., Wolfe, E. C., DeWs’s, 55’. D., and Hotfnian, F. A. Serum folate and vitamin B 1 2 levels in P.iti#{128}’ntswith small elI lung

cancer.

K. D., Gore, H., Mueller, H., and Krumdieck, dysplasia associated with folic acid therapy Am. I. Clin. Nutr., 35: 73-82, 1982.

R., (Thu. I.. Mandelson, invasive cervical cancer.

6. Ziegler, R. R. S., Mallin, Hoover, R. N. women in the 7.

L. Constitutive 1 199-1204,

Diagnostic Cytology I. B. Lippincott Co.,

.1. Butterworth, (T. Improvement in users of oral

I.

14. laiob, F., CoIiian, N., radioassav for two vit,imins: 30:616, 1977.

Acknowledgments

(

1 1. Longo. folate.

19. Cavanagh, cervix: some 2 36, 1981.

SI:

306-

310,

Nutrition

and

1984. [)iet

Therapy.

0., Praphat, H. and Ruffolo, current views. Obstet. and

20. Butterworth, dysplasia. Am.

C. E., and I. Cliii. Nutr.,

Monterey,

E. H. Car Gynecology

Norris, D. Folic acid :17: 332 -333, 1983.

and

CA:

inoma of the uterine Annual, 10: 1 93 vitamin

C in cervical

a

2 1 . Krunidieck, (1 sensitive Med. Genet.,

(T. L., ,in(J Howard-Peebles, P. N. On the nature fragile sites in human c hromosonies: an hypothesis. 16: 23 28, 1983.

22. Goulian,

M., Bleile,

inucirporation 1960, 1980.

of uracil

23. W.

Heimburger, in smokers 1988.

DNA.

0). C., Alexander,

C., and Krunidieck,

plasia 1530,

B., and Tseng, mId)

treated

Proc.

with

tolate

and

of Iiili Aiii. I.

B. Y. Methotrexate-induecl NatI.

B., Birch,

C. L. Improvement

Wads-

Acad.

Sci.

USA,

R., Buttervvorth,

in bronchial vitamin

Bt2.

urns77:

1 956-

C. E., Bailey,

squamous JAMA,

259:

meta1525-

24.

Lashner, B. A., 1-leidenreich, P. A., Sn, C. I. ., Kane, S. V., and Hanauer, S. B. Effect of folale supplementation on the inc idence of clyspl,isis md canc er in chronic ulc erative colitis. Gastroenterology, 97: 255-259, 1989 25. Wassertheil-Smoller, S., Romne, S. L., \Vylie-Rosett, I., Slagle, Miller, C., Lucido, D., Duttagupta, C., and Palan, P. R. I)ietarv vitamin and uterine cervical dysplasia. Am. I. Epidemiol.. I 14: 7 14 724, 1981. 26. Riininey, S. F., Duttagupta. C., S., Dwyer, A., Wassertheil-Snioller, nun C and uterine cervid al clvsplasia. 980, 1985. 27. Glatthaar, B. F., Hornig, acid in carcinogenesis. Adv.

S., C

Basu, I., Palan, P. R., Karp, S.. SI.mgle, S., And Wylie-Rosett, I. Plasma vit,iAm. I. Obstet. Gv’nrnoI., I 1 S 976

D. H., ,mncl Moser, Li. 1 he role if ,is i)rl)i( Exp. Med. Biol.. 206: 357 377. 1986.