Carcinogenesis vol.16 no.2 pp.

223-23O, 1995

A comparison of lymphocyte micronuclei and plasma micronutrients in vegetarians and non-vegetarians

Michael Fenech1 and Josephine Rinaldi

CSIRO Division of Human Nutrition, PO Box 10041, Gouger Street, Adelaide, SA 5000, Australia 'To whom correspondence should be addressed

We performed a biochemical and cytogenetic epidemiological study to establish if there are significant differences between vegetarians (V) and non-vegetarians (NV) in their peripheral blood lymphocyte micronucleus (MN) index, which is a measure of chromosome damage rate. The levels of plasma vitamin C (VIT-C), vitamin E (VIT-E), vitamin B12 (B12) and folic acid were also analysed to assess if differences in chromosome damage rates were associated with these potentially antimutagenic micronutrients. Volunteers were classified as either 'vegetarian' if they had abstained from eating any flesh foods for at least 3 years prior to the study or 'non-vegetarian' if they consumed meat or meat products at least 5 days/week for at least 3 years before participation in the study. The volunteers in the study consisted of 47 male and 79 female V and 66 male and 72 female NV, all of whom were non-smokers for at least 3 years prior to the study. The age of the volunteers varied between 20 and 89 years. There was no significant difference in the slope of the age-related increase in MN index of V and NV of either sex. However, the MN index was significantly lower in NV males in the age group 2040 years and significantly lower for V males in the 41-60 years age group. No difference between the MN index of older males was detectable and there also was no difference in the MN index of V and NV females across all age groups. V were generally found to have significantly higher plasma levels of VIT-C and folic acid, significantly lower levels of B12, and similar levels of VIT-E when compared with NV. VIT-C correlated positively with MN index in young males, but the reverse was true for B12. In young females folate and B12 appeared to correlate negatively with MN index. VIT-E had no apparent impact on MN index. These data suggest that the level of folate and B12 may be more important than VIT-C or VIT-E in minimizing chromosome damage rates in human lymphocytes. Overall, the data from this study do not support the hypothesis that V have a lower genetic damage rate than NV.

Introduction Several studies have established that the development of cancer involves the accumulation of several mutations at critical genes starting at initiation through to progression and metastasis ( 1 5). The rate at which cancer develops is dependent on germline
Abbreviations: SCE, sister chromatid exchange; V, vegetarian; NV, nonvegetarian; CBMN assay, cytokinesis-block micronucleus assay; MN, micronucleus; VIT-C, vitamin C; VIT-E, vitamin E; B12, vitamin B l 2 ; PHA, phytohaemagglutinin; BN, binucleated cell; MNi, micronuclei; CHOL, cholesterol; Mf, MN frequency in females; Mm, MN frequency in males. Oxford University Press

mutations that are inherited, as well as spontaneous mutations occurring as a result of endogenous or exogenous genotoxic agents. It is logical to assume that a decline in genetic damage rate may postpone the onset of cancer and it is, therefore, interesting to determine if genetic damage rates are reduced in those populations that are likely to experience a lower cancer rate due to genetic or lifestyle factors. It has been established that in cancer-prone syndromes, such as ataxia telangiectasia and Bloom's syndrome, the spontaneous chromosome damage rates are abnormally elevated (6-8). Conversely, one could expect that in populations, such as the ovo-lacto vegetarian Seventh Day Adventists, for which there is a documented lowered cancer rate for specific cancers when compared with omnivores (9,10), one might expect that mutation rates are relatively low. In view of the emerging interest in vegetarianism (11), it is important to establish if such a lifestyle is in fact linked with lowered genetic damage rates. To the best of our knowledge there have been only two relatively small studies comparing sister chromatid exchanges (SCEs*) in vegetarians (V) and non-vegetarians (NV) and both of these involved Seventh Day Adventists (12,13). These studies showed that SCEs in lymphocytes of children of Seventh Day Adventists were similar to those of children from the general population, but the SCE level in adult members (mean age 65 years) of this religious group was significantly lower (by 30%) than that of the general population. Interpretation of these studies was limited by the numbers of subjects and lack of verification of nutrient intake by biochemical analysis. In view of our ongoing research on the impact of diet and other environmental factors on chromosome damage rate assayed using the cytokinesis-block micronucleus (CBMN) assay (14-16), we were interested in: (i) determining if the micronucleus (MN) frequencies in lymphocytes of V and NV are different; (ii) assessing if the levels of micronutrients that may be expected to influence chromosome damage rate are different in V and NV; and (iii) establishing if any differences in MN index could be accounted for by differences in antimutagenic micronutrients. The micronutrients analysed were the antioxidants ascorbic acid (VIT-C) and a-tocopherol (VIT-E) and the B vitamins folic acid and vitamin B| 2 (B12). Information from such a study should be important because: (i) it would help establish the extent to which differences in diet could confound genetic damage studies in human populations; (ii) micronutrients that could significantly modify chromosome damage rates in vivo may be identified; (iii) it may help determine if 'normal' micronutrient intakes are optimal for minimizing chromosome damage rate; and (iv) it may provide the baseline data required for the appropriate design of intervention studies investigating the antimutagenic effects of specific micronutrients. Materials and methods
Volunteer recruitment Volunteers were recruited by placing an advertisment in local newspapers and by direct contact with various local vegetarian and vegan societies. The

223

M.Fenech and J.Rinaldi recruited volunteers consisted of 150 females and 114 males aged between 20 and 89 years, with a minimum of 15 volunteers/decade for both males and females. A total of 126 volunteers were classified as V on the basis that they had abstained from eating any flesh foods (red meat, chicken or fish) for at least 3 years prior to the study and 138 were classified as NV on the basis that they consumed meat or meat products at least 5 days/week for at least 3 years before joining the study. All volunteers who were current smokers, smokers who had ceased smoking less than 3 years before the study, diagnosed with cancer or had regular complicating usage of medicine were excluded from the study. Questionnaires The questionnaire for details on lifestyle, occupation and health status was identical to the one published by Carrano and Natarajan (17) and the dietary questionnaire used was developed at the CSIRO Division of Human Nutrition (18). Detailed analysis of dietary data from these questionnaires and their relation to MN frequency data will be reported separately. Blood collection and CBMN assay To avoid possible confounding by dietary metabolites and diurnal effects, all volunteers participating in this study were required to fast overnight and to donate a blood sample between 8.00 and 11.00 a.m. before having breakfast, which was provided. Between 10 and 30 ml blood was collected in heparinized plastic tubes. Lymphocytes were isolated on Ficoll-Hypaque gradients and the standard CBMN assay was employed as previously described (14,19). Briefly, lymphocytes were cultured in plastic round-bottom tubes at a concentration of lX10 6 /ml in McCoy's 5A medium, stimulated to divide with phytohaemagglutinin (PHA) (Wellcome) and 44 h later 4.5 |lg/ml cytochalasin-B (Sigma) were added to accumulate cells that had divided once only as binucleated cells (BN). Cells were then harvested by cytocentrifugation at 72 h post-PHA stimulation, air-dried, fixed in methanol and stained using Diff-Quik (LAB-AIDS, Australia). Micronuclei (MNi) were scored in 1000 BN according to published criteria (14). The same person (J.R.) performed the CBMN assay and scored the slides for the whole of this study to minimize variation due to operator differences. Furthermore, slides were coded so that the operator was not aware of the age, sex and dietary habits of the donors when scoring the slides. Analysis of plasma micronulrients Fresh plasma was obtained from the heparinized blood and snap frozen in liquid nitrogen. Each sample was analysed in duplicate for the four micronutrients using the following techniques: B12 and folic acid were measured simultaneously using radioimmunoassay (BioRad, Quantaphase II); VIT-C was measured spectrophotometrically using the dinitrophenylhydrazine method of Roe and Kuether (20); and VIT-E was measured using HPLC by the method of Hatam and Kayden (21). Because VIT-E is lipid soluble, it is often reported as a ratio with cholesterol (CHOL) concentration. We have therefore also measured CHOL, using a COBAS analyser and an in vitro diagnostic kit (Roche Products) and computed the VIT-E:CHOL ratio. Statistical analysis of data Descriptive statistics, comparisons between groups, correlations and regression analysis were performed using the procedures described in Instat (GraphPad Software Inc.) and CSS Statistica (StatSoft). Regression analysis of the relationship between MN frequency and age was performed using the least squares method; for regression and correlation analysis missing data were case-wise deleted. Initial data analysis of plasma micronutrients (16) indicated that there were weak, but significant, correlations between micronutrient levels and between micronutrient status and age. To avoid the potential confounding effects of collinearity when using multiple regression analysis, we first age-adjusted (16) and then log-transformed the age-adjusted MN frequency of each

MALE NON-VEGETARIANS

MALE VEGETARIANS

100

3
8
o o o

r=0.564 P<0.0001

slope=0.302 Y int=1.36
80

r=0.505 P<0.0003

slope= 0.268 Y int= 2.38

80

60

o o o
o

60

40

o o

o a o
20
40 60 AGE IN YEARS

20
Oi

o
"tr o o
20

Vj
o 80

"b
100

80

100

40 60 AGE IN YEARS

FEMALE NON-VEGETARIANS

FEMALE VEGETARIANS 100

100
C O

CRONLJCLEI/'1000 BN CEI

80
m

r=0.509 P<0.0001

slope=0.482 Y int=1.93

r=0.615 P<0.0001

slope=0.502 Y int= -1.04

80

o o o
LJ

60

60

40

.-'

*M?'*^'^

40

20
20

. .

20

40 60 AGE IN YEARS

80

100

40 60 AGE IN YEARS

Fig. 1. Scattergrams and linear regression analysis of MN frequency in relation to age in V and NV males and V and NV females. BN, cytokinesis-blocked binucleated cells; int, intercept.

224

Micronuclei, micronutrients and vegetarianism individual and then determined the regression weight of each micronutrient independently using simple regression analysis. The non-parametric Mann-Whitney U-test was used for comparison between groups using the data for each individual within a group; MN data were age- and sex-adjusted prior to analysis to minimize the confounding effect of these important variables (14,15).

Results Preliminary analysis of the pooled data showed that MN frequency in females (Mf) was significantly higher than the MN frequency in males (Mm) in all decades examined (P < 0.05), the Mf:Mm ratio varied between 1.47 and 1.65 (mean SEM 1.53 0.03) and showed no trend with age. As expected from previous studies, the MN index increased with age (Figure 1) and the rate of this increase was 0.482 MNi7 1000 BN/annum for NV females and 0.502 MNi/1000 BN/ annum for V females. The rate of increase for NV males was 0.302 MNi/1000 BN/annum and for V males 0.268 MNi/1000 BN/annum. The differences between the slopes for V and NV of the same sex were not statistically significant. In view of the importance of sex and age as variables

influencing the MN index, we analysed the data for females and males separately and in each case comparisons for MNi and micronutrients were performed after the individuals were subdivided into three age groups, namely 20-40, 41-60 and 61-90 years. The MN data for each individual in each age group were adjusted to the median age of each group (i.e. 30, 50 and 75 years respectively) using the slopes of the agerelated regression lines illustrated in Figure 1. The MN index in 20-40-year-old males was significantly lower in NV when compared with V, but in the 41-60-yearold males the reverse was true, i.e. the MN index was significantly higher in NV males (Figure 2A). In the 61-90year-old males there were no significant differences between V and NV. In females there were no significant differences in the MN index of V and NV in all of the age groups examined, however, there was a trend for higher MN frequency in NV in the 41-60-year-old group (Figure 2B). Plasma B12 values in V males were on average lower than that for NV males (Figure 3A). The differences between V and NV males were most pronounced in the 20-40-year-old group, in which the mean plasma B12 of V was 160.2 pmol/1 and that of NV was 352.4 pmol/1. The extent of this difference

(A)
MN FREQUENCY IN VEG. AND NONVEG. MALES 40 20-40Y
C/> _J _l LJ O CO

(A)
PLASMA B12 IN VEG. AND NONVEG. MALES

35 30

41-60Y

61-90Y

N=16

N=24

25 20 15 10 5 0
N=15 N=18 *

8
o cr
X
1, | Q.

V NV

V NV

V NV

500 450 400 350 300 250 200 150 100 50 0

20-40y

41-60y

61-90y

(B)
MN FREQUENCY IN VEG. AND NONVEG. FEMALES 40 20-40Y 35 30 o 25 41-60Y 61-90Y

NV

NV

V NV

(B)
PLASMA B12 IN VEG. AND NONVEG. FEMALES
400 350 ^
o

20-40y

41-60y
n=27 n=23 ~]~

61-90y

300
n=29
|

8 20
a. 15 Id a. 10 5 0

" 250 S 200

PLASMA

V NV

V NV

V NV

50
0

Fig. 2. Comparison of MN frequency in (A) V and NV males and (B) V and NV females. To facilitate comparison of data between sexes, data for females were sex-adjusted to male values using the 1.53 factor described in the results section. *P = 0.024 (one-tailed test); **P = 0.027. Significance was determined using the Mann-Whitney one-tailed test.

NV

NV

V NV

Fig. 3. Comparison of plasma B12 in (A) V and NV males, *P = 0.008, **P = 0.018, and (B) V and NV females, *P = 0.013. Significance was determined using the MannWhitney two-tailed test.

225

M.Fenech and J.Rinaldi

between V and NV declined with age, mainly due to a progressively lower plasma B12 in older NV, possibly due to reduced meat intake or reduced absorption of this micronutrient. Plasma B12 status in females (Figure 3B) was somewhat different, in that plasma B12 in V females was consistently higher than the values observed for their V male counterparts and, unlike the males, the plasma B12 status showed a positive trend with age. A significantly lower plasma B12 was observed in V females relative to NV females only in the 20-40-yearold age group; the difference was comparably smaller than that observed between V and NV males of the same age group (i.e. a 44% difference between V and NV females as compared with a 120% difference between V and NV males). In the 2040-year-old group, 16.4% of the V females, 1.3% of the NV females, 25.0% of the V males and 9.0% of the NV males had plasma B12 values below 110 pmol/1, which is the highest value within the deficient range usually recorded in patients showing clinical symptoms of pernicious anaemia (22). Plasma folate levels were consistently higher in V males and females relative to NV males and females in all age groups, with statistically significant increments in 20-40-yearold V males and females, and 41-60- and 61-90-year-old V

females (Figure 4A and B). The highest mean plasma folate level was 44.5 nmol/1, recorded in 41-60-year-old V females, and the lowest mean level was 21.0 nmol/1, measured in 2040-year-old NV males. None of the plasma values in all the individuals studied were below 3.4 nmol/1, which is the highest level in the range usually recorded for anaemic patients as diagnosed by a range of haematological and biochemical tests (22). Plasma VIT-C levels were consistently higher in V males and V females relative to their NV counterparts across all age groups (Figure 5A and B). These differences were statistically significant (P < 0.05) in all groups studied with the exception of the 61-90-year-old female group, in which P = 0.065. There were no obvious alterations in plasma VIT-C levels with age in both sexes in both V and NV. A plasma level of 1.5 mg/100 ml usually indicates saturation (19); 79 and 75% of V males and females respectively and 50 and 47% of NV males and females respectively had plasma levels of VIT-C above 1.5 mg/100 ml. Plasma VIT-E:CHOL ratios did not vary significantly between V and NV across all age groups in both sexes (Figure 6A and B).

(A)
PLASMA FOLATE IN VEG. AND NONVEG. MALES

(A)
PLSMA V I T - C IN MALE VEG. AND NONVEG.

V V NV V NV V NV

NV

V NV

NV

(B)
PLASMA FOLATE IN VEG. AND NONVEG. FEMALES
3.0

(B)
PL/WA V I T - C IN VEG AND NONVEG. FEMALES

20-40y

41-60y
n=27' *

61-90y

i
o ^ V

25
n=23 n=29

2.0 1.5
1.0 05 0.0
n=20

T
n=29

SB
"-I-.

NV

NV

NV

V NV

V NV

V NV
Fig. 5. Comparison of plasma VIT-C in (A) V and NV males, *P = 0.009, **P = 0.019, ***P = 0.007, and (B) V and NV females, *P = 0.010, **P = 0.0001. Significance was determined using the Mann-Whitney two-tailed test.

Fig. 4. Comparison of plasma folate in (A) V and NV males, *P = 0.035, and (B) V and NV females, *P = 0.034, **P = 0.0001, ***P = 0.022. Significance was determined using the MannWhitney two-tailed test.

226

Micronuclei, micronutrients and vegetarianism

A separate analysis of this data set, in which data from all age groups were grouped together, suggested that plasma VITC was the only micronutrient showing a significant correlation with MN frequency; the correlation was positive and occurred in males, but not in females (16; Figure 7). We therefore decided to examine whether the statistically significant differences in the MN frequencies observed in V and NV males could be explained by their plasma micronutrient status. We examined the correlation between each plasma micronutrient and MN frequency in the 20-40-year-old group and the 41-60-yearold group after combining the data for V and NV (Table I). This analysis showed that: (i) there was no significant correlation between plasma micronutrient status and MN frequency in the 41-60-year-old group; and (ii) that the MN frequency in the 20-40-year-old group was, however, correlated positively with plasma folate (r = 0.453), plasma VIT-E:CHOL (r = 0.366) and plasma VIT-C (r = 0.388) and correlated negatively with plasma B12 (r = -0.203). The interpretation of the latter correlations is confounded by relatively strong correlations between plasma folate and plasma VIT-E:CHOL (r = 0.465, P < 0.011) and plasma folate and plasma VIT-C (r = 0.526, P < 0.003). The apparent increasing impact of plasma

micronutrients on MN frequency with decreasing age is also emphasized by a separate analysis of the 20-30-year-old males in this population sample, which shows a stronger positive correlation between plasma VIT-C and MN frequency (r = 0.823) and a strong negative correlation between plasma B12 and MN frequency (r = 0.799); interpretation of these results is also confounded by the strong negative correlation between plasma B12 and plasma VIT-C (r = -0.772, P < 0.002). Analysis of data for females did not reveal any significant correlations between plasma micronutrients and MN frequency in the specific age groups for which MN frequency differences were examined, however, separate data analysis for 20-30year-old females in this population sample revealed a significant negative correlation between MN frequency and combined plasma folate and plasma B12 (r = -0.463, P < 0.030) (16). Discussion Some epidemiological studies have shown that the incidence of specific cancers and overall mortality rate is reduced in populations who abstain from or minimize their intake of flesh foods (9,23,24). However, to date there is no clear explanation for such an observation, although several hypotheses have been suggested, such as reduced caloric intake (25), higher intake of antioxidant vitamins such as VIT-C and p-carotene (26) and reduced intake of potent mutagens found in cooked meat (27). The somatic mutation theory of cancer suggests that genetic damage rates should be reduced in those individuals or groups who have a lower risk for cancer. In view of the possibility that vegetarianism may indeed be associated with a lowered genetic damage rate (12) we decided to perform a more detailed study using the MN index in lymphocytes, which is an established biomarker for chromosome breakage and chromosome loss. The CBMN assay used in this study is now widely adopted for population monitoring due to its relative ease of implementation (14). For example, it was practical for one laboratory to screen the MN index of up to 900 Japanese

(A)
PLASMA V I T - E / C H O L . RATIO IN VEG. AND N0NVEG. MALES

a. o

i
o V NV V NV V NV

REGRESSION WEIGHTS OF PLASMA MICRONUTRIENTS IN RELATION TO L0G lo [AGE-ADJUSTED MN FREQUENCY] i FEMALE

(B)
PLASMA VIT-E/CHOLESTEROL RATIO IN FEMALE VEG AND NONVEG.

o
LJ

1
d
cr

6
4

20-40y
n=20 n=29

41-60y

61-90y
n=23 1~ n=29
|

~r :-;

T ggg:

a. o
ui
CL

32r 1

it

VIT-E/CHOL

VIT-C

NV

NV

NV

Fig. 6. Comparison of plasma VIT-E:CH0L ratio in (A) V and NV males and (B) V and NV females.

Fig. 7. Regression weights of plasma micronutrients in relation to log-transformed, age-adjusted MN frequency. Regression weights were determined by analysing the effect of each plasma micronutrient independently of other micronutrients. The MN frequency of each subject was adjusted to their expected value if they were 55 years old using the equation MN55y = (55 - A) S + M, where MN55y is MN frequency adjusted to the expected value if subject was 55 years old, A is actual age in years, M is actual MN frequency and 5 is the slope of the regression line for MN frequency in relation to age. **P = 0.014,

227

M.Fenech and J.Rinaldi

Table I. Correlation between MN frequency and plasma micronutrients in males Age group (years) n Correlation coefficients FOL 20-30 20-40 41-60 13 29 40 0.392 (P -- 0.185) = 0.453 (P == 0.013) = 0.185 (P = 0.254) B12 -0.799 (P = 0.001) -0.203 (P = 0.291) 0.174 (P = 0.284) VIT-E:CHOL ratio -0.184 (P = 0.547) 0.366 (P = 0.050) 0.100 (P = 0.537) VIT-C 0.823 (P = 0.001) 0.388 (P = 0.037) 0.128 (P = 0.430)

atomic bomb survivors to establish if there is a link between radiosensitivity and survival (28). Because the MN assay has an important role as an effective biological dosimeter of exposure to ionizing radiation (14,29) and exposure to chemical genotoxins (30), it is important to ascertain which lifestyle and dietary factors may have a significant impact on this index in order to be able to correctly interpret differences observed between individuals and/or populations. To date we have identified age and sex as being the most important variables affecting the MN index and this has been confirmed by other laboratories (14,15,28). MN frequency is -1.5 times higher in females relative to males, which could be due to the loss of the X chromosome. The molecular changes affecting loss of X chromosomes are not well understood and although undercondensation of the inactive X chromosome has been implicated as being the most probable mechanism (31), it is still not clear if altered nutritional status (e.g. methyl donor deficiency) could alter the rate of X chromosome loss in females. It is, therefore, probable that the comparison of MN frequencies between V and NV females may have been confounded by the random loss of X chromosomes, which would have made the detection of small differences in genetic damage between groups more difficult to achieve. This study also identifies variation in the MN index that could be attributable to dietary differences. These dietary differences were not only ascertained by dietary questionnaires, but also by analysis of several micronutrients in plasma that could be expected to differ, to varying extents, in V and NV volunteers. The most important point to emerge concerning the MN frequencies is that the overall difference between V and NV subjects, in both males and females, was not remarkable; when differences were apparent (i.e. males aged 20-40 and 41-60 years) they did not invariably suggest a lower genetic damage rate in V subjects. One possible reason why no remarkable differences were observed in this study is the likelihood that our NV volunteers included a large proportion of individuals who led a healthy lifestyle and who ate a balanced and nutritious diet that effectively only differed from that of the V volunteers by the extent of meat intake as a protein source; this possibility is being explored by ongoing detailed analysis of the dietary questionairres submitted by the volunteers. Nevertheless, there were clear and statistically significant differences in plasma micronutrients between V and NV that were characterized by relatively low B12 and relatively high folate and VIT-C in V relative to NV. These differences go in the expected directions and indicate a proportionately higher fruit and vegetable intake in V relative to NV (32,33). Both folate and B12 play an important role in the maintenance of genetic integrity and gene expression by their central roles in the synthesis of deoxythymidine monophosphate and the maintenance of methylation of 5-methylcytosine; deficiencies in these micronutrients are known to elevate chromosome breakage rate and MN frequency (34-44). While
228

considerable focus has been placed on the importance of folate supplementation as a means of protection against cancer (4547) and the incidence of neural tube defects (48,49), little emphasis has been made of the role of B12, even though deficiency in this micronutrient is common, especially in V individuals. Recent evidence suggests that: (i) plasma folate and B12 are independent risk factors for neural tube defects; and (ii) that increased risk occurs at levels of folate and B12 that are not associated with clinical manifestation of deficiency (50). Together, these findings indicate that both folate and B12 intake should be optimal to minimize pathological effects. Clearly, neither the strictly vegetarian nor the omnivorous diet with inadequate fruit and vegetable intake are optimal in this regard. B12 and folate were the only micronutrients in this study that showed any significant negative correlation with MN frequency, suggesting that the accepted normal values for plasma may not be optimal for minimizing genetic damage rates. Considerable controversy exists on the potential role of VITC in protection against cancer. Clearly a high intake of fresh fruit and vegetables is associated with a high VIT-C and folic acid intake, but it is not clear if these micronutrients are, in fact, the main cause of the associated lower cancer rates. While some studies show protection of VIT-C against cancer (51-55), others suggest that high VIT-C aggravates cancer risk in chemically induced animal cancers (56,57) and there is good evidence for its clastogenic action in vitro (55,58,59) and in vivo (60). Our studies indicate a clear positive correlation between MN frequency and plasma VIT-C in males, but do not prove that VIT-C itself is the direct cause. This could only be ascertained by intervention studies with high doses of VITC. Of all the micronutrients examined, VIT-E appears to have the least influence on the MN index, suggesting either that lipid peroxidation is unlikely to have important secondary effects on DNA integrity or that the range of VIT-E values in this study was too narrow to discern any dose effects. Some recent epidemiological studies have suggested that protection against cancer and coronary heart disease by VIT-E was only evident in those individuals who supplemented their normal dietary intake with VIT-E capsules (6163). In analysing the relationship of plasma micronutrients and MN frequency in males it became evident that the influence of micronutrient status was increasingly pronounced with decreasing age, to the extent that the plasma micronutrients examined had no obvious impact on MN frequency in age groups above 40 years, while significant and progressively higher correlations could be detected in age groups below 40 years. The correlations observed have been somewhat confounded by other significant correlations between the plasma levels of the micronutrients examined. These findings suggest that the reduced MN frequency in V males in the 4 1 60 year age group could not be explained by the differences in micronutrient status between V and NV, but the reverse was true with regard to the MN frequency reduction in NV

Micronuclei, micronutrients and vegetarianism

males in the 20-30 and 20-40 year age groups. These data are indicative of a possible role for VIT-C in aggravating DNA damage rates in young males and a possible beneficial effect associated with elevated B12 status. The data for females in the 20-30 year age group also show a dose-related reduction in MN frequency in relation to the combined folate and B12 plasma levels, thus reinforcing the notion that genetic damage rates in lymphocytes may be minimized by optimizing folate and B12 intake beyond the currently accepted normal plasma values. Our observations do not appear to support the view that lowered cancer rates in vegetarians could be explained by their higher VIT-C intake nor do they suggest that VIT-C supplementation would be a successful approach to minimize genetic damage rates in lymphocytes. Because MN frequency appeared to be effectively influenced by micronutrient status only in the younger age groups, it seems that dietary interventions investigating the relationship between diet and genetic damage would be most successful if restricted to individuals aged less than 30 years. This appears to make sense, because it would be expected that agerelated biological changes, such as reduced absorption of micronutrients and increased genetic damage rate, are likely to overwhelm any doseresponse effects produced by normal intake of miconutrients in the older categories; this effectively also implies that supplementation over and above normal dietary intake may be required to observe any effects on genetic damage rates in the elderly. In conclusion, more attention needs to be given to the influence of diet on baseline genetic damage rates because: (i) it helps to identify those micronutrients and those age groups with which positive effects are most likely to be obtained; and (ii) it enables a more refined interpretation of observed genetic damage rates when the MN assay is used as a biomarker of exposure to physical or chemical genotoxins. The lack of an exceptional difference in spontaneous genetic damage in lymphocytes of V and NV does not exclude the possibility that mutation rates in other tissues may be different. Acknowledgements
We would like to acknowledge the contribution of the numerous volunteers who participated in this study, as well as the help of Dr Peter Clifton, Sr Rosemary McArthur and Sr Maria Nugent, who assisted us in the collection of blood. In particular we would also like to acknowledge the supporting technical assistance provided by Clare Aitken at various critical points of the project and Dr Ivor Dreosti for reviewing the manuscript. This research was partially funded by financial grants from the Anti-Cancer Foundation of South Australia and the Australian Meat Research Corporation.

References
l.Loeb.L.A. (1991) Mutator phenotype may be required for multistage carcinogenesis. Cancer Res., 51, 3075-3079. 2. Yuspa,S.H. and Poirier.M.C. (1988) Chemical carcinogenesis: from animal models to molecular models in one decade. Adv. Cancer Res., 50, 25-71. 3.Suzuki,K., Suzuki,F., Watanabe.M. and Nikaido.O. (1989) Multi-step nature of X-ray-induced neoplastic transformation in golden hamster embryo cells: expression of tranformed phenotypes and stepwise changes in karyotypes. Cancer Res., 49, 2134-2140. 4.Mukhtar,H. and Bickers.D.R. (1993) Environmental skin cancer: mechanisms, models and human relevancemeeting report. Cancer Res., 53, 3439-3442. 5. Fearon.E.R. and Vogelstein.B. (1990) A genetic model for colorectal tumorigenesis. Cell, 61, 759-767. 6.Knight,R.D., Parshad.R., Price.F.M., Tarone.R.E. and Sanford.K.K. (1993) X-ray-induced chromatid damage in relation to DNA repair and cancer incidence in family members. Int. J. Cancer, 54, 589-593. 7.McKinnon,P.J. (1987) Ataxia-telangiectasia: an inherited disorder of ionising-radiation sensitivity in man. Hum. Genet., 75, 197-208.

8. Rosin.M.P. and Ochs.H.D. (1986) In vivo chromosome instability in ataxia telangiectasia homozygotes and heterozygotes. Hum. Genet., 74, 335-340. 9. Fraser.G.E., Beeson.W.L. and Phillips.R.L. (1991) Diet and lung cancer in California Seventh-day Adventists. Am. J. Epidemiol, 133, 683-693 10. Phillips.R.L. and Snowdon.D.A. (1983) Association of meat and coffee use with cancers of the large bowel, breast and prostate among SeventhDay Adventists. Cancer Res., 43, 2403s-2408s. 11. American Dietetic Association (1993) Position of the American Dietetic Association: vegetarian diets. J. Am. Dietet. Ass., 93, 1317-1319 12.Wulf,H.C, Iversen,A.S., Husum,B. and Niebuhr.E. (1986) Very low sister chromatid exchange rate in Seventh-Day Adventists. Mutat. Res., 162, 131-135. 13.Hermansen,R., Waksvik,H. and Fonnebo.V. (1991) Sister chromatid exchange in children of Seventh-Day Adventists and matched controls. Carcinogenesis, 12, 423425. 14. Fenech,M. (1993) The cytokinesis-block micronucleus technique: a detailed description of the method and its application to genotoxicity studies in human poulations. Mutat. Res., 285, 35-44. 15.Fenech,M., Neville.S. and RinaldiJ. (1994) Sex is an important variable affecting spontaneous micronucleus frequency in cytokinesis-blocked lymphocytes. Mutat. Res., 313, 203-207. 16. Fenech,M. and RinaldiJ. (1994) The relationship between micronuclei in human lymphocytes and plasma levels of vitamin-C, vitamin-E, vitamin B12 and folic acid. Carcinogenesis, 15, 1405-1411. 17.Carrano,A.V. and Natarajan,A.T. (1988) Considerations for population monitoring. Mutat. Res., 204, 379-406. 18.Baghurst,K.I. and Record.S.J. (1984) A computerised dietary analysis system for use with diet diaries of food frequency questionnaires. Community Hlth Studies, 7, 11-18. 19. Fenech.M. and Morley.A.A. (1985) Measurement of micronuclei in lymphocytes. Mutat. Res., 147, 29-36. 20. Roe,J.H. and Kuether.C.A. (1942) The determination of ascorbic acid in blood and urine through the 2,4-dinitrophenylhydrazine derivative of dehydroascorbic acid. J. Biol. Chem., 147, 399^107. 21.Hatam,L.J. and Kayden,H.J. (1979) A high-performance liquid chromatographic method for the determination of tocopherol in plasma and cellular elements of the blood. J. Lipid Res., 20, 639-645. 22. Freidrich.W. (1988) Vitamins. Walter de Gruyter, Berlin, Germany. 23.Armstrong,B. and Doll.R. (1975) Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int. J. Cancer, 15, 617-631. 24. Gerhardsson de Verdier.M., Hagman.U., Steineck.G., Reiger.A. and Norell.S.E. (1990) Diet, body mass and colorectal cancer: a case-referent study. Int. J. Cancer, 46, 832-838. 25.Lutz,W.K. and SchlatterJ. (1992) Chemical carcinogens and overnutrition in diet-related cancer. Carcinogenesis, 13, 2211-2216. 26. Smith.A.H. and Waller.K.D. (1991) Serum beta-carotene in persons with cancer and their immediate families. Am. J. Epidemiol., 133, 661-671. 27. Gerhardsson de Verdier.M., Hagman.U., Peters.R.K., Steineck.G. and Overick.E. (1991) Meat, cooking methods and colorectal cancer: a casereferent study in Stockholm. Int. J. Cancer, 49, 520-525. 28.Ban,S., Cologne.J.B., Fujita.S. and Awa.A.A. (1993) Radiosensitivity of atomic bomb survivors as determined with a micronucleus assay. Radiat. Res., 134, 170-178. 29. Fenech.M., Denham,J., Francis,W. and Morley,A.A. (1990) Micronuclei in cytokinesis-blocked lymphocytes of cancer patients following fractionated partial-body radiotherapy. Int. J. Radiat. Biol., 57, 373-383. 30. Osanto.S., Thijssen.J.C.P., Woldering.V.M., van Rijn,J.L.S., Natarajan.A.T. and Tates.A.D. (1991) Increased frequency of chromosomal damage in peripheral blood lymphocytes up to nine years following curative chemotherapy of patients with testicular carcinoma. Environ. Mol. Mutagen., 17, 71-78. 31. Guttenbach.M. and Schmid.M. (1994) Exclusion of specific human chromosomes into micronuclei by 5-azacytidine treatment of human lymphocytes. Exp. Cell Res., 211, 127-132. 32. Subar.A.F., Block,G. and James.L.D. (1989) Folate intake and food sources in the U.S. population. Am. J. Clin. Nutr., 50, 508-516. 33.Block,G., Dresser.C.M., Hartmann.A.M. and Carroll.M.D. (1985) Nutrient sources in the American diet: quantitative data from the NHANES D " survey. Part I: vitamins and minerals. Am. J. Epidemiol., 122, 13-26. 34. MacGregorJ.T. (1990) Dietary factors affecting spontaneous chromosomal damage in man. In Aeschbacher.H.U. (ed.), Mulagens and Carcinogens in the Diet. Wiley-Liss, New York, pp. 139-153. 35.James,S.J. and Yin,L. (1989) Diet-induced DNA damage and altered nucleotide metabolism in lymphocytes from methyl-donor-deficient rats. Carcinogenesis, 10, 1209-1214. 36. Castro,C.E. (1987) Nutrient effects on DNA and chromatin structure. Ann.

229

M.Fenech and J.Rinaldi Rev. JVw/r., 7, 407^*21. 37.Menzies,R.C, Crossen,P.E., Fitzgerald,P.H. and Gunz,F.W. (1966) Cytogenetic and chemical studies on marrow cells in B12 and folate deficiency. Blood, 28, 581-594. 38.Rana,S.R., Colman,N., Goh.K.O., Herbert,V. and Klemperer.M.R. (1983) Transcobalamin II deficiency associated with unusual marrow findings and chromosomal abnormalities. Am. J. Hemalol., 14, 89-96. 39.Everson,R.B., Wehr.C.M., Erexson.G.L. and MacGregorJ.T. (1988) Association of marginal folate depletion with increased human chromosomal damage in vivo: demonstration by analysis of micronucleated erythrocytes. / Natl Cancer Inst., 80, 525-529. 4O.Chen,A.T.L., ReidyJ.A., Annest.J.L., Welty.T.K. and Zhou.H. (1989) Increased chromosome fragility as a consequence of blood folate levels, smoking status, and coffee consumption. Environ. Mol. Mutagen., 13, 319-324. 41.Jacky,P.B., Beek.B. and Sutherland,G.R. (1983) Fragile sites in chromosomes: possible model for the study of spontaneous chromosome breakage. Science, 220, 69-70. 42. Heath.C.W. (1966) Cytogenetic observations in vitamin B12 and folate deficiency. Blood, 27, 800-815. 43.Libbus,B.L., Borman.L.S., Ventrone.C.H. and Branda.R.F. (1990) Nutritional folate deficiency in CHO cells: chromosomal abnormalities associated with perturbations in nucleic acid precursors. Cancer Genet. Cytogenet., 46, 231-242. 44.Branda,R.F., O'NeillJ.P, Sullivan.L.M. and Albertini,R.J. (1991) Factors influencing mutation at the HPRT locus in T-lymphocytes: women treated for breast cancer. Cancer Res., 51, 6603-6607. 45. Krumdieck.C.L. (1991) Localised folate deficiency and cancer. In Laidlaw.S.A. and Swendsied.M.E. (eds), Vitamins and Cancer Prevention. Wiley-Liss, New York, pp. 39-50. 46. Giovannucci.E., Stampfer,M.J., Colditz.G.A., Rimm,E.B., Trichopoulos.D., Rosner,B.A., Spiezer.F.E. and Willet.W.C. (1993) Folate, methionine and alcohol intake and risk of colorectal adenoma. J. Natl Cancer Inst., 85, 875-883. 47.Butterworth,C.E.,Jr, Hatch,K.D., Macaluso,M., Cole,P., Sauberlich.H., Soong.S., Borst,M. and Baker,V.V. (1992) Folate deficiency and cervical dysplasia. J. Am. Med. Assoc, 267, 528-533. 48. Densem.J., Frost.C. and Stone,R. MRC Vitamin Study Research Group. (1991) Prevention of neural tube defects: results of the Medical Research Council Vitamin Study. Lancet, 338, 130-137. 49.Czeizel,A.E. and Dudas.I. (1992) Prevention of the first occurrence of neural tube defects by periconceptional vitamin supplementation. New Engl. J. Med., 327, 1832-1835. 50.Kirke,P.N., Molloy.A.M., Daly.L.E., Burke.H., Weir.D.G. and Scott.J.M. (1993) Maternal plasma folate and vitamin Bj 2 are independent risk factors for neural tube defects. Q. J. Med., 86, 703-708. 51.Sarma,L. and Kesavan.P.C. (1993) Protective effects of vitamins C and E against gamma-ray-induced chromosomal damage in mouse. Int. J. Radiat. Biol, 63, 759-764. 52.Rossner,P, Cerna,M, Pokorna,D., Hajek.V. and Petr.J. (1988) Effect of ascorbic acid prophylaxis on the frequency of chromosome aberrations, urine mutagenicity and nucleolus test in workers occupationally exposed to cytostatic drugs. Mutat. Res., 208, 149-153. 53. Sram,R.J., Dobias.L., Pastorkova.A., Rossner,P. and Janca.L. (1983) Effect of ascorbic acid prophylaxis on the frequency of chromosome aberrations in the peripheral lymphocytes of coal-tar workers. Mutat. Res., 120, 181-186. 54. Fraga,C.G., Motchnik.P.A., Shigenaga,M.K., Helbock,H.J., Jacob,R.A. and Ames.B.N. (1991) Ascorbic acid protects against endogenous oxidative DNA damage in human sperm. Proc. Natl Acad. Sci. USA, 88, 1100311006. 55. Shamberger,R.J. (1984) Genetic toxicology of ascorbic acid. Mutat. Res., 133, 135-159. 56. Shklar.G., SchwartzJ., Trickler.D. and Cheverie.S.R. (1993) The effectiveness of a mixture of beta-carotene, alpha-tocopherol, glutathione and ascorbic acid for cancer prevention. Nutr. Cancer, 20, 145-151. 57.Wooley,P.V., Kumar.S., Fitzgerald.P. and Simpson.R.T. (1987) Ascorbate potentiates DNA damage by 1 -methyl-1-nitrosourea in vivo and generates DNA strand breaks in vivo. Carcinogenesis, 8, 1657-1662. 58.Bruchelt,G., Schraufstatter.I.U., Niethammer.D. and Cochrane,C.G. (1991) Ascorbic acid enhances the effects of 6-hydroxydopamine and H 2 O 2 on iron-dependent DNA strand breaks and related processes in the neuroblastoma cell line SH-NSH. Cancer Res., 51, 6066-6072. 59. Stich,H.F., Wei.L. and Whiting.R.F. (1980) Chromosome aberrations in mammalian cells exposed to vitamin C and multiple vitamin pills. Fd Cosmet. Toxicol., 18, 497-501. 60. Shelby,M.D., Erexson,G.L., Hook,G.J. and Tice,R.R. (1993) Evaluation of a three-exposure mouse bone-marrow micronucleus protocol: results with 49 chemicals. Environ. Mol. Mutagen., 21, 16-179. 61.Rimm,E.B., Stampfer.M.J., Ascherio.A., Giovannucci.E., Colditz,G.A. and Willet,W.C. (1993) Vitamin E consumption and the risk of coronary heart disease in men. New Engl. J. Med, 328, 1450-1456. 62. Stampfer.M.J., Hennekens,C.H., MansonJ.E., Colditz.G.A., Rosner,B. and Willet,W.C. (1993) Vitamin E consumption and the risk of coronary disease in women. New Engl. J. Med., 328, 1444-1449. 63.Bostick,R.M., Potter.J.D., McKenzie.D.R., Sellers.T.A., Kushi.L.H., Steinmetz,K.A. and Folsom.A.R. (1993) Reduced risk of colon cancer with high intake of vitamin E: the Iowa Women's Health Study. Cancer Res., 53, 4230-4237. Received on August 16, 1994; revised on October 24, 1994; accepted on November 7, 1994

230