- Email: [email protected]
The effect of smoking on placental and fetal zinc status
The effect of smoking on placental and fetal zinc status P.M. Kuhnert, Ph.D., B. R. Kuhnert, Ph.D., P. Erhard, B.S., W. T. Brashear, M.S., S. :L Groh-Wargo, M.S., and S. Webster, M.S. Cleveland, Ohio Previous studies have reported a cadmium/zinc interaction in cadmium-exposed pregnant animals that results in (1) increased placental cadmium levels, (2) increased placental zinc levels, and (3) decreased placental zinc transport. This study was carried out to determine whether zinc status would be affected in pregnant women exposed to cadmium through cigarette smoke. Atomic absorption spectroscopy was used to determine the levels of cadmium and zinc; 65 pregnant women who smoke and 84 who do not smoke were studied. Our data reveal that increased cadmium levels in pregnant women as the result of smoking increase placental zinc levels and decrease cord red blood cell zinc levels. Significantly higher levels of both cadmium and zinc were found in the placentas of pregnant women who smoke; moreover, stepwise multiple regression showed that maternal whole blood cadmium levels predicted placental zinc levels. In regard to cord blood, a significant 9% decrease in the red blood cell zinc level was observed in infants of mothers who smoke and this decrease was correlated with smoking activity, as evaluated by measuring plasma levels of thiocyanate. Also cord red blood cell zinc levels were found to correlate with placental zinc levels in nonsmokers but not in smokers. Overall, our data show that a cadmium/zinc interaction does take place in the maternal-fetal-placental unit of pregnant women who smoke and results in less favorable zinc status in the infants. (AM J 0BSTET GYNECOL 1987;157:1241-6.)
Key words: Smoking, placental zinc, fetal zinc
Zinc nutritional status has a profound influence on the biochemistry and toxicology of cadmium. One reason for this is that cadmium and zinc often act antagonistically in biologic systems; for example, cadmium acts as an antagonist to zinc in zinc-requiring metalloenzymes, for instance, carbonic anhydrase,' alkaline phosphates; and others.' Another reason is that metallothionein, which is produced in response to cadmium exposure, binds both cadmium and zinc; thus, if zinc levels are marginally adequate, zinc status may be aggravated by the binding and storage of zinc by cadmium-induced metallothionein. Finally, an inadequate zinc nutritional status can influence the absorption and tissue uptake of cadmium, thereby increasing the probability of a toxic effect due to cadmium.' Pregnant women may be particularly susceptible to
From the Departments of Obstetrics and Gynecology and Nutrition Services, Cleveland Metropolitan General Hospital, and the Perinatal Clinical Research Center of Case Western Reserve University. Supported in part by National Institutes of Health Gra>!ts 5M01RR-00210 and R01-HD-17015, United States Public Health Service. Presented in part at the Thirty-third Annual Meeting of the Society for Gynecologic Investigation, Toronto, Ontario, Canada, March 19-22, 1986, and at the Seventieth Annual Meeting of the Federation of American Societies of Experimental Biology, St. Louis, Missouri, April13-18, 1986. Received for publication October 3, 1986; revised May 22, 1987; accepted june 3, 1987. Reprint requests: B. R. Kuhnert, Ph.D., Department of Obstetrics and Gynecology, 3395 Scranton Road, Cleveland, OH 44109.
this cadmium/zinc interaction. They are susceptible because their zinc intake is generally substantially lower than the Recommended Daily Allowance of zinc during pregnancy" and significant associations have been reported between fetomaternal complications and plasma zinc levels." Even more vulnerable, however, is the pregnant woman who smokes cigarettes. The average cigarette contains as much as 1 1-1g of cadmium, and heavy smokers may have an intake of 20 1-1g/day or more as the result of smoking, in addition to the average daily intake of 50 1-1g/day from dietary sources. Thus decreased zinc nutritional status and increased cadmium intake may promote a cadmium/zinc interaction in pregnant women who smoke. The magnitude of this cadmium/zinc interaction in pregnant women who smoke is of course important since it could influence fetal growth and development. Low zinc concentrations in maternal blood and cord blood have been related to abnormal events in parturition, intrauterine growth retardation, and increased incidences of congenital malformations in humans. 5 Moreover, previous research on pregnant animals has clearly shown that this interaction occurs between cadmium and zinc and can result in the fetus receiving less zinc.' Accordingly, we investigated the zinc status and cadmium levels of pregnant women who smoke. Our primary purpose was to determine whether the cadmium levels were related to zinc status in the mother and fetus. We report herein the results of our investigation.
1241
1242
Kuhnert et al.
November 1987 Am J Obstet Gynecol
Table I. Characteristics of the maternal study population Smokers (n = 65) Characteristic
Maternal age (yr) Race Black White Other Parity Nulliparous Multiparous Delivery method Vaginal Cesarean section Primary Repeat Complications None Hypertension Rh sensitization Preeclampsia Diabetes Class A Classes B to F and R Mild anemia Weight >200 pounds <100 pounds Precipitate labor Long Jabot Prolonged second stage Premature rupture of membranes
n
I
%
24 ± 4
Table II. Characteristics of the infant study population Nonsmokers (n = 84)
I
n
%
.25 ± 4
I9 45 I
29 69 2
34 44 6
40 52* 7
I2 53
18 82
I6 68
19 81
35
54
40
48
9 21
14 32
33
II
13 39
43 I 0 2
66 2 0 3
53
63
2
2
2
I
2 3
4
5
4
6
3
4
5 0 5 2 I
8 0 8 3 2
12 0 7 3 4
14 0 8 4 5
0
0
0
0
I I
I
I I
Smoking mother (n = 65) Characteristic
Birth weight (gm) Gestation (wk) Apgar score <7 I min 5 min Sex Male Female Complications Small for dates Large for dates Premature Postterm
n
I
Nonsmoking mother (n = 84)
%
316I ± 597 38.8 ± 2
n
1
%
3515 ± 533* 39.3 ± 2
2
17 3
II
37 28
57 43
43 41
51 49
5 10
8 15
0 32
0* 39
4
6 2
5 5
6 6
II
I
I
13 I
*p < 0.05.
I
*p < 0.05.
Methods and material
Patients and collection of samples. Sixty-five pregnant women who smoke and 84 who do not smoke and their fetuses were included in this study. The only selection criteria were that the mothers resided in or very near Cleveland, Ohio, and were delivered of their infants at Cleveland Metropolitan General Hospital. They were not selected with regard to prenatal or in· trapartum medical problems, although in depth clinical information was recorded. Informed consent for this study was obtained from the subjects in the appropriate manner. Samples of blood were obtained from the mothers within an hour of delivery and from the umbilical cord immediately after delivery. Samples of 10 ml were procured in separate heparinized polypropylene syringes from the antecubital vein of the mother and umbilical cord vein of the infant. Samples of placental tissue were obtained after several cotyledons had been perfused with 2 L of cold 0.9% saline solution to remove excess blood, and the membranes and blood vessels were re-
moved. Each sample was then stored frozen until analyzed for levels of whole blood cadmium (maternal blood only), placental cadmium, placental zinc, plasma zinc, and red blood cell zinc. Maternal plasma was also analyzed for thiocyanate, which was used in cm·~unc tion with questionnaires for assessing smoking behavior. Zinc intake in a random subgroup of the population (29 smokers, 37 nonsmokers) was assessed by diet history at term. The diet history included a 7-day food frequency assessment that was verified for accuracy by a 24-hour recall. Food sources known to be the major contributors of zinc in the diet (for example, meat, seafood, eggs, dairy products, dried beans and peas, and whole grain breads and cereals) were quantified and an estimate of weekly zinc intake in milligrams was calculated. Determination of cadmium, zinc, and thiocyanate levels. The samples of blood and placental tissue were prepared for analysis of cadmium by means of acid digestion. The digestion mixture was concentrated Ultrex (J. T. Baker, Phillipsburg, New Jersey) nitric acid and hydrogen peroxide. The details of this procedure were previously described by us. 7 An atomic absorption spectrophotometer (Model 560, Perkin Elmer Corp., Norwalk, Connecticut) equipped with a deuterium background corrector and a graphite furnace (Model HGA-2200, Perkin Elmer Corp.) was used for the cadmium analyses. In contrast to the analysis of cadmium, flame atomic absorption analysis was used for the analysis of zinc. Red blood cell and plasma zinc was analyzed after a 1 : 4 dilution with deionized water. Glycerol was added
Placental and fetal zinc status
Volume 157 Number 5
1243
Table III. Indices of smoking Smokers (N = 65) (mean± SD)
Thiocyanate (IJ.mol!L) Maternal whole blood cadmium (ng/gm) Placental cadmium (ng/gm)
107 ± 40 1.33 ± 0.8 12.0 ± 7.5
Nonsmokers (N = 84) (mean± SD)
30 0.80 8.1
± ± ±
14 0.4 5.0
Significance
p < 0.05 p < 0.05 p < 0.05
Table IV. Indices·of zinc status
Mother Serum alkaline phosphatase (units/100 ml) Plasma zinc (j.Lg/100 ml) Red blood cell zinc (fLg/100 ml) Placenta Zinc (ng/gm) Alkaline phosphatase (units/100 ml) Infant cord blood Alkaline phosphatase (units/100 ml) Plasma zinc (j.Lg/ I 00 ml) Red blood cell zinc (ng/gm)
to the blank and standards to approximately match the viscosity of the diluted plasma. Placental zinc was analyzed after acid digestion as described under cadmium analysis. Plasma thiocyanate, a byproduct of the metabolism of cyanide in cigarette smoke, was measured by means of the procedure described by Pettigrew and Fell." Patients with thiocyanate values ,;59 J-lmol/L were defined as nonsmokers and those with values of ~60 J-lmol/L were defined as smokers in the few cases where information obtained from the clinical record was not consistent with the thiocyanate data. Passive smokers were considered nonsmokers unless the thiocyanate values were ~60. Maternal, fetal, and placental alkaline phosphatase levels were quantitated with the method of Doellgast and Meis." Data analysis. Statistical analyses were performed with the BMDP UCLA statistical software package and a DEC Vax 111750 computer (Digital Equipment Corporation, Maynard, Massachusetts) with a Unix operating system. Data analysis consisted of descriptive statistics, t tests, and simple correlations to produce an overview of significant differences between the groups and interrelationships between the various biochemical indices. The X2 analysis was used to test for differences in clinical parameters between the two groups. Then stepwise multiple regression techniques were used to examine the interactions between cadmium exposure and zinc status while controlling for statistically confounding variables such as maternal age, parity, race, delivery method, vitamin intake, and diabetes. In some cases (placental cadmium, placental zinc, thiocyanate),
Suwkrn (N = 65)
Nonsmokers (N = 84)
Significance
6.3 ± 2.8 57.5 ± 9.7 1232 ± 169
6.0 ± 2.3 57.0 ± 10.0 1218 ± 150
NS NS NS
12.1 ± 2.7 9.0 ± 4.5
11.1 ± 2.8 7.4 ± 3.6
p < 0.05 p < 0.05
9.4 :t 2.8 81.1 :t 14.5 230 ± 55
10.0 ± 3.1 83.2 ± 15.0 250 ± 60
NS NS p < 0.05
the variables were not normally distributed. In these cases, log transformations were used. Statistical significance was the same regardless of whether a log transformation was used. Significance was accepted at p < 0.05. Results
The characteristics of the mothers and infants in this study are shown in Tables I and II, respectively. The smokers and nonsmokers are classified separately. Table I shows that the maternal population groups had similar characteristics except for the proportiOn of white patients who smoked; a significantly higher proportion of the white patients were smokers. Table II shows that the mean birth weight of the infants of mothers who smoked was less than that of the infants of mothers who did not smoke (t = 4.7; p < 0.0001). Also more small-for-dates infants were born to mothers who smoked than to mothers who did not smoke. These findings are in agreement with previous studies concerning pregnant women who smoke. The blood levels of thiocyanate of the two study populations are shown in Table III. The mean levels in the mothers who smoked were significantly higher than those in the nonsmokers. Illustrated in Fig. 1 and Table III are the relative levels of cadmium found in maternal blood and in the placentas of smokers and nonsmokers. The percent increase in cadmium in the whole blood of smokers was 64%; the percent increase in placental cadmium was 45%. In both smokers and nonsmokers more cadmium was found in the placentas than in the blood. These
1244
Kuhnert et al.
14
rn
.,0 .... a.
-o
D
November 1987 Am .J Obstet Gynecol
0
10
10
't: G
.,
=
•,.
u
•c
::J
6
6
E
"'I
N
'"'c
-
ii
3
3:
..
..
'-
G
Ill
...
Smokers
~ Non-Smokers
0
j:
280
14
Smokers Non-Smokers
240
c
N
u II a:
.
220
::J
"' 200 2
2
0
Smoker• Whole Blood Cd
Placental Cd
Placental Zn
Fig. 1. The effect of smoking on whole blood cadmium, placental cadmium, and placental zinc levels.
findings are in agreement with previous studies concerning cadmium exposure in pregnant women who smoke."· 10' 12 The zinc status of smokers and nonsmokers and that of their infants (as assessed by measurements of plasma and red blood cell zinc levels and alkaline phosphatase activity) is compared in Table IV. No significant differences were observed in the mothers. Moreover, diet histories verified that there was no difference in zinc intake between pregnant women who smoked and those who did not. The average daily intake of zinc was 10.4 mg. A major finding of the study was the 9% decrease in the red blood cell zinc level in the cord blood of infants of mothers who smoke (Fig. 2); this difference is statistically significant. A comparison of the zinc status of the mothers and infants shows that the infants had plasma zinc and alkaline phosphatase levels that were approximately one third higher than those of their mothers; on the other hand, the red blood cell zinc levels of the infants were approximately five times lower than those observed in the mothers. Comparisons of the levels of zinc in the placentas of smokers and nonsmokers are shown in Table IV and Fig. 1. Significantly higher levels of zinc were found in the placentas of pregnant women who smoked. The significantly higher levels of placental zinc correlated with significantly higher levels of alkaline phosphatase in the placenta. In smokers, the significant predictor of both placental zinc and placental cadmium by stepwise multiple regression was maternal whole blood cadmium. In nonsmokers, the maternal whole blood cadmium level pre-
NonSmoker•
Fig. 2. Cord vein red blood cell zinc levels in smokers and nonsmokers.
dieted the placental cadmium level but not the placental zinc level. These data are summarized in Table V, which shows the F value, degrees of freedom, and p value for each analysis. Since a 9% decrease in the red blood cell Zn level was observed in the cords of infants of mothers who smoke, the relationships between the biochemical variables and cord vein red blood cell zinc levels were examined. Significant predictors of the cord vein red blood cell zinc level, as assessed by stepwise multiple regression, were different in the smoking and nonsmoking groups. In smokers, thiocyanate was the strongest predictor of the cord vein red blood cell zinc level. The simple correlation showed that the relationship was negative, that is, the more thiocyanate in maternal plasma, the less zinc in the red blood cells of the infant. In nonsmokers, placental zinc was the strongest predictor of cord vein red blood cell zinc. The simple correlation showed that the relationship was positive, that is, the more placental zinc, the more cord vein red blood cell zinc. These data are summarized in Table VI, which shows the F value, degrees of freedom, p value, and r value for each analysis.
Comment The principal purpose of this study was to determine whether the increased levels of cadmium found in pregnant women who smoke affects the distribution of zinc in the maternal-fetal-placental unit. Our data show that increased cadmium levels as the result of smoking influence the levels of zinc in the placenta and cord blood and that this effect of cadmium is significant even though the absolute increase in cadmium as the result of smoking is small. To determine the effect of increased cadmium ex-
Volume 157 Number 5
Placental and fetal zinc status
1245
Table V. Significant relationships between maternal whole blood cadmium and placental cadmium or zinc in smokers and nonsmokers obtained with stepwise multiple regression Smokers Placental cadmium Placental zinc
F[2,59] F[l,60]
= =
Nonsmokers
4.6; p < 0.05 15.3; p < 0.001
F[l,83] = 4.7; p < 0.05 F[2,82] = 3.0; NS
Table VI. Predictors and univariate correlation coefficients of cord vein red blood cell zinc in smokers and nonsmokers Predictor Stepwise multiple regression Univariate correlation coefficient
Smokers
Nonsmokers
Thiocyanate F[2,58] = 4.14; p < 0.05 r = 0.21; p < 0.05
Placental zinc F[2,82] = 5.95; p < 0.025 r = 0.21; p < 0.05
posure on zinc status in pregnant women who smoke, we first verified that our study population of pregnant smokers and nonsmokers (with similar clinical characteristics) had similar levels of zinc intake. This was done by comparing diet histories and several indices of zinc status of the two groups. Our data showed that there was no difference in zinc intake between the smokers and the nonsmokers and that there was no difference in plasma zinc and red blood cell zinc levels. The data also showed that zinc intake of both groups was about 50% below the Recommended Daily Allowance for zinc during pregnancy, as reported by other investigators studying healthy pregnant patients. 5 However, since both groups had decreased zinc intake and were similar in all other respects, a comparison could be made in regard to the effect of cadmium exposure on zinc status. The cadmium exposure of our study population as the result of smoking was similar to that reported by us 6 and by others. 10"12 We observed a 64% increase in whole blood cadmium and a 45% increase in placental cadmium in the present study. The percent increase in placental cadmium reported to date has ranged from 26 10 to 400. 11. 12 The values differ, of course, depending on the study population and on the analytic techniques used. However, all the studies agree that there is a substantial increase in cadmium in the placentas of pregnant women who smoke. Increased exposure to cadmium and decreased zinc intake, as found in our pregnant smokers, are conditions that favor an effect of smoking on zinc status. Data supporting this are based on studies involving animals. For example, studies by Webb and Samarawickrama13 on rats show that maternal cadmium exposure causes fetal zinc deficiency and that this is one mechanism of the teratogenic effects of cadmium. By giving cadmium to pregnant rats these workers were able to show a dose-related decrease in fetal uptake of
zinc. Additional evidence that cadmium can induce fetal zinc deficiency was reported by Daston. 11 This author showed that coadministration of zinc almost eliminated the severe lesions in the fetal lung when the pregnant rat was given cadmium. Although the levels of cadmium in these studies are much higher than we observed in our study population, the threshold for such effects has not been established for animals or humans. Moreover, our data suggest that a cadmium/zinc interaction takes place in the maternal-fetalplacental unit of pregnant women who smoke. One aspect of our data that suggests a cadmium/zinc interaction and an effect of smoking on zinc status is the 10% increase in the level of zinc in the placentas of pregnant women who smoke. This is suggestive because it is likely that the increased levels of cadmium as the result of smoking induce the production of metallothionein, a low molecular weight protein that binds both cadmium and zinc. With the increased production of metallothionein there is an increased binding of cadmium and zinc and consequently a sequestering of these metals in the placenta. Although this has not been specifically studied in the placenta, autopsy studies of the kidney have shown that zinc levels increase in the kidney cortex with increasing concentrations of cadmium, and this increase is almost equimolar to the increase in cadmium. 15 Support for this interpretation is also provided by our statistical analysis, which shows that in smokers maternal whole blood cadmium levels are predictive not only of the placental cadmium levels but also of the placental zinc levels. Another aspect of our data that suggests an effect of smoking on zinc status is the 9% decrease in red blood cell zinc observed in the cords of infants of mothers who smoke. This decrease was found to correlate with the levels of thiocyanate (an index of the number of cigarettes smoked) in maternal blood. Thus the greater the number of cigarettes smoked, the lower the level
1246
Kuhnert et al.
of red blood cell zinc observed in cord blood. An explanation for this, which is supported by our data, is that zinc bound to metallothionein is accumulating in the placenta and that this binding may reduce the amount of zinc available to the fetus. Another possibility is that less zinc is being transported by the placenta to the fetus; this has been shown to occur in cadmiuminjected pregnant rats. 16 Regardless of the explanation, however, the data show that a cadmium/zinc interaction occurring in pregnant women who smoke may be harmful to the fetus since the decrease in cord red blood cell zinc is related to decreased birth weight in the infants of smokers. 17 Finally, additional evidence of the effect of smoking on fetal zinc status was noted in our data when the relationship between placental zinc levels and cord vein red blood cell zinc levels was examined. The data show a positive correlation between placental zinc and cord vein red blood cell zinc in nonsmokers; in smokers, however, this relationship was not found. The cadmium/zinc interaction, that is, the altered metabolism of zinc brought about by increased cadmium intake due to cigarette smoking, may explain the lack of this correlation in smokers. The effect of smoking on zinc status in the placenta and fetus may have no health significance if the availability of zinc is above the real requirement. On the other hand, this interaction may be detrimental to the fetus where zinc status is marginal. Additional studies are needed to clarify this point and to determine if zinc supplementation can counteract these changes in tissue levels (as it does in animals). Nevertheless, our studies show that there is an adverse effect of smoking on zinc status in the placenta and fetus, that is, a cadmium/zinc interaction does take place in the maternal-fetalplacental unit of pregnant women who smoke. This finding provides one additional reason for discouraging pregnant women from smoking. REFERENCES I. Lindskog S, Malmstrom BG. Metal binding and catalytic
activity in bovine carbonic anhydrase. J Bioi Chern 1962;237:1129.
November 1987 Am .J Obstet Gynecol
2. Plocke DJ, Levinthal C, Vallee BL. Alkaline phosphatase of Escherichia coli: a zinc metalloenzyme. Biochemistry 1962; 1:373. 3. Vallee BL. Biochemistry, physiology, and pathology of zinc. Physiol Rev 1959;39:443. 4. Parzyck DC, Shaw SM, Kessler MV, Vetter RJ, van Sickle DC, Mayes RA. Fetal effects of cadmium in pregnant rats on normal and zinc-deficient diets. Bull Environ Contam Toxicol 1978; 19:206. 5. Apgar]. Zinc and reproduction. Ann Rev Nutr 1985; 5:43. 6. Kuhnert PM, Kuhnert BR, Bottoms SF, Erhard P. Cadmium levels in maternal blood, fetal cord blood, and placental tissues of pregnant women who smoke. AM J OBSTET GYNECOL 1982;142:1021. 7. Kuhnert PM, Erhard P, Kuhnert BR. Analysis of cadmium in whole blood and placental tissues by furnace atomic absorption spectroscopy. In: Hemphill DD, ed. Trace substances in environmental health. Columbia, Missouri: University of Missouri Press, 1982 vol 16:370. 8. Pettigrew AR, Fell GS. Simplified colorimetric determination of thiocyanate in biological fluids, and its application to investigation of the toxic amblyopias. Clin Chern 1972; 18:996. 9. Doellgast GJ, Meis PJ. Use of specific inhibitors to discriminate alkaline phosphatase isoenzymes originating from human liver, placenta and intestine: absence of meconial alkaline phosphatase in maternal serum. Clin Chern Acta 1979;25: 1230. 10. Roels H, Hubermont G, BuchetJP, Lauwerys R. Placental transfer of lead, mercury, cadmium, and carbon monoxide in women. Environ Res 1978; 16:236. 11. Miller RK, Gardner KA. Cadmium in the human placenta: relationship to smoking. Teratology 1981 ;23:51. 12. Miller RK, Kellogg CK. The pharmacodynamics of prenatal chemical exposure. Bethesda, Maryland: National Institute on Drug Abuse, 1985:39. (Research monographs; series 60). 13. Webb M, Samarawickrama GP. Placental transport and embryonic utilization of essential metabolites in the rat at the teratogenic dose of cadmium. J Appl Toxicol 1981; 1,5:270. 14. Daston GP. Fetal zinc deficiency as a mechanism for cadmium-induced toxicity to the developing rat lung and pulmonary surfactant. Toxicology 1982;24:55. 15. Nordberg M, Elinder CG, Rahnster B. Cadmium, zinc, and copper in horse kidney metallothionein. Environ Res 1979;20:341. 16. Samarawickrama GP, Webb M. In: Proceedings of international conference on cadmium. Bethesda, Maryland: National Institutes of Health, 1978. 17. Kuhnert BR, Kuhnert PM, Debanne S, Williams TG. The relationship between cadmium, zinc, and birth weight in pregnant women who smoke. AM J 0BSTET GYNECOL 1987;157:1247.
Key words: Smoking, placental zinc, fetal zinc
Zinc nutritional status has a profound influence on the biochemistry and toxicology of cadmium. One reason for this is that cadmium and zinc often act antagonistically in biologic systems; for example, cadmium acts as an antagonist to zinc in zinc-requiring metalloenzymes, for instance, carbonic anhydrase,' alkaline phosphates; and others.' Another reason is that metallothionein, which is produced in response to cadmium exposure, binds both cadmium and zinc; thus, if zinc levels are marginally adequate, zinc status may be aggravated by the binding and storage of zinc by cadmium-induced metallothionein. Finally, an inadequate zinc nutritional status can influence the absorption and tissue uptake of cadmium, thereby increasing the probability of a toxic effect due to cadmium.' Pregnant women may be particularly susceptible to
From the Departments of Obstetrics and Gynecology and Nutrition Services, Cleveland Metropolitan General Hospital, and the Perinatal Clinical Research Center of Case Western Reserve University. Supported in part by National Institutes of Health Gra>!ts 5M01RR-00210 and R01-HD-17015, United States Public Health Service. Presented in part at the Thirty-third Annual Meeting of the Society for Gynecologic Investigation, Toronto, Ontario, Canada, March 19-22, 1986, and at the Seventieth Annual Meeting of the Federation of American Societies of Experimental Biology, St. Louis, Missouri, April13-18, 1986. Received for publication October 3, 1986; revised May 22, 1987; accepted june 3, 1987. Reprint requests: B. R. Kuhnert, Ph.D., Department of Obstetrics and Gynecology, 3395 Scranton Road, Cleveland, OH 44109.
this cadmium/zinc interaction. They are susceptible because their zinc intake is generally substantially lower than the Recommended Daily Allowance of zinc during pregnancy" and significant associations have been reported between fetomaternal complications and plasma zinc levels." Even more vulnerable, however, is the pregnant woman who smokes cigarettes. The average cigarette contains as much as 1 1-1g of cadmium, and heavy smokers may have an intake of 20 1-1g/day or more as the result of smoking, in addition to the average daily intake of 50 1-1g/day from dietary sources. Thus decreased zinc nutritional status and increased cadmium intake may promote a cadmium/zinc interaction in pregnant women who smoke. The magnitude of this cadmium/zinc interaction in pregnant women who smoke is of course important since it could influence fetal growth and development. Low zinc concentrations in maternal blood and cord blood have been related to abnormal events in parturition, intrauterine growth retardation, and increased incidences of congenital malformations in humans. 5 Moreover, previous research on pregnant animals has clearly shown that this interaction occurs between cadmium and zinc and can result in the fetus receiving less zinc.' Accordingly, we investigated the zinc status and cadmium levels of pregnant women who smoke. Our primary purpose was to determine whether the cadmium levels were related to zinc status in the mother and fetus. We report herein the results of our investigation.
1241
1242
Kuhnert et al.
November 1987 Am J Obstet Gynecol
Table I. Characteristics of the maternal study population Smokers (n = 65) Characteristic
Maternal age (yr) Race Black White Other Parity Nulliparous Multiparous Delivery method Vaginal Cesarean section Primary Repeat Complications None Hypertension Rh sensitization Preeclampsia Diabetes Class A Classes B to F and R Mild anemia Weight >200 pounds <100 pounds Precipitate labor Long Jabot Prolonged second stage Premature rupture of membranes
n
I
%
24 ± 4
Table II. Characteristics of the infant study population Nonsmokers (n = 84)
I
n
%
.25 ± 4
I9 45 I
29 69 2
34 44 6
40 52* 7
I2 53
18 82
I6 68
19 81
35
54
40
48
9 21
14 32
33
II
13 39
43 I 0 2
66 2 0 3
53
63
2
2
2
I
2 3
4
5
4
6
3
4
5 0 5 2 I
8 0 8 3 2
12 0 7 3 4
14 0 8 4 5
0
0
0
0
I I
I
I I
Smoking mother (n = 65) Characteristic
Birth weight (gm) Gestation (wk) Apgar score <7 I min 5 min Sex Male Female Complications Small for dates Large for dates Premature Postterm
n
I
Nonsmoking mother (n = 84)
%
316I ± 597 38.8 ± 2
n
1
%
3515 ± 533* 39.3 ± 2
2
17 3
II
37 28
57 43
43 41
51 49
5 10
8 15
0 32
0* 39
4
6 2
5 5
6 6
II
I
I
13 I
*p < 0.05.
I
*p < 0.05.
Methods and material
Patients and collection of samples. Sixty-five pregnant women who smoke and 84 who do not smoke and their fetuses were included in this study. The only selection criteria were that the mothers resided in or very near Cleveland, Ohio, and were delivered of their infants at Cleveland Metropolitan General Hospital. They were not selected with regard to prenatal or in· trapartum medical problems, although in depth clinical information was recorded. Informed consent for this study was obtained from the subjects in the appropriate manner. Samples of blood were obtained from the mothers within an hour of delivery and from the umbilical cord immediately after delivery. Samples of 10 ml were procured in separate heparinized polypropylene syringes from the antecubital vein of the mother and umbilical cord vein of the infant. Samples of placental tissue were obtained after several cotyledons had been perfused with 2 L of cold 0.9% saline solution to remove excess blood, and the membranes and blood vessels were re-
moved. Each sample was then stored frozen until analyzed for levels of whole blood cadmium (maternal blood only), placental cadmium, placental zinc, plasma zinc, and red blood cell zinc. Maternal plasma was also analyzed for thiocyanate, which was used in cm·~unc tion with questionnaires for assessing smoking behavior. Zinc intake in a random subgroup of the population (29 smokers, 37 nonsmokers) was assessed by diet history at term. The diet history included a 7-day food frequency assessment that was verified for accuracy by a 24-hour recall. Food sources known to be the major contributors of zinc in the diet (for example, meat, seafood, eggs, dairy products, dried beans and peas, and whole grain breads and cereals) were quantified and an estimate of weekly zinc intake in milligrams was calculated. Determination of cadmium, zinc, and thiocyanate levels. The samples of blood and placental tissue were prepared for analysis of cadmium by means of acid digestion. The digestion mixture was concentrated Ultrex (J. T. Baker, Phillipsburg, New Jersey) nitric acid and hydrogen peroxide. The details of this procedure were previously described by us. 7 An atomic absorption spectrophotometer (Model 560, Perkin Elmer Corp., Norwalk, Connecticut) equipped with a deuterium background corrector and a graphite furnace (Model HGA-2200, Perkin Elmer Corp.) was used for the cadmium analyses. In contrast to the analysis of cadmium, flame atomic absorption analysis was used for the analysis of zinc. Red blood cell and plasma zinc was analyzed after a 1 : 4 dilution with deionized water. Glycerol was added
Placental and fetal zinc status
Volume 157 Number 5
1243
Table III. Indices of smoking Smokers (N = 65) (mean± SD)
Thiocyanate (IJ.mol!L) Maternal whole blood cadmium (ng/gm) Placental cadmium (ng/gm)
107 ± 40 1.33 ± 0.8 12.0 ± 7.5
Nonsmokers (N = 84) (mean± SD)
30 0.80 8.1
± ± ±
14 0.4 5.0
Significance
p < 0.05 p < 0.05 p < 0.05
Table IV. Indices·of zinc status
Mother Serum alkaline phosphatase (units/100 ml) Plasma zinc (j.Lg/100 ml) Red blood cell zinc (fLg/100 ml) Placenta Zinc (ng/gm) Alkaline phosphatase (units/100 ml) Infant cord blood Alkaline phosphatase (units/100 ml) Plasma zinc (j.Lg/ I 00 ml) Red blood cell zinc (ng/gm)
to the blank and standards to approximately match the viscosity of the diluted plasma. Placental zinc was analyzed after acid digestion as described under cadmium analysis. Plasma thiocyanate, a byproduct of the metabolism of cyanide in cigarette smoke, was measured by means of the procedure described by Pettigrew and Fell." Patients with thiocyanate values ,;59 J-lmol/L were defined as nonsmokers and those with values of ~60 J-lmol/L were defined as smokers in the few cases where information obtained from the clinical record was not consistent with the thiocyanate data. Passive smokers were considered nonsmokers unless the thiocyanate values were ~60. Maternal, fetal, and placental alkaline phosphatase levels were quantitated with the method of Doellgast and Meis." Data analysis. Statistical analyses were performed with the BMDP UCLA statistical software package and a DEC Vax 111750 computer (Digital Equipment Corporation, Maynard, Massachusetts) with a Unix operating system. Data analysis consisted of descriptive statistics, t tests, and simple correlations to produce an overview of significant differences between the groups and interrelationships between the various biochemical indices. The X2 analysis was used to test for differences in clinical parameters between the two groups. Then stepwise multiple regression techniques were used to examine the interactions between cadmium exposure and zinc status while controlling for statistically confounding variables such as maternal age, parity, race, delivery method, vitamin intake, and diabetes. In some cases (placental cadmium, placental zinc, thiocyanate),
Suwkrn (N = 65)
Nonsmokers (N = 84)
Significance
6.3 ± 2.8 57.5 ± 9.7 1232 ± 169
6.0 ± 2.3 57.0 ± 10.0 1218 ± 150
NS NS NS
12.1 ± 2.7 9.0 ± 4.5
11.1 ± 2.8 7.4 ± 3.6
p < 0.05 p < 0.05
9.4 :t 2.8 81.1 :t 14.5 230 ± 55
10.0 ± 3.1 83.2 ± 15.0 250 ± 60
NS NS p < 0.05
the variables were not normally distributed. In these cases, log transformations were used. Statistical significance was the same regardless of whether a log transformation was used. Significance was accepted at p < 0.05. Results
The characteristics of the mothers and infants in this study are shown in Tables I and II, respectively. The smokers and nonsmokers are classified separately. Table I shows that the maternal population groups had similar characteristics except for the proportiOn of white patients who smoked; a significantly higher proportion of the white patients were smokers. Table II shows that the mean birth weight of the infants of mothers who smoked was less than that of the infants of mothers who did not smoke (t = 4.7; p < 0.0001). Also more small-for-dates infants were born to mothers who smoked than to mothers who did not smoke. These findings are in agreement with previous studies concerning pregnant women who smoke. The blood levels of thiocyanate of the two study populations are shown in Table III. The mean levels in the mothers who smoked were significantly higher than those in the nonsmokers. Illustrated in Fig. 1 and Table III are the relative levels of cadmium found in maternal blood and in the placentas of smokers and nonsmokers. The percent increase in cadmium in the whole blood of smokers was 64%; the percent increase in placental cadmium was 45%. In both smokers and nonsmokers more cadmium was found in the placentas than in the blood. These
1244
Kuhnert et al.
14
rn
.,0 .... a.
-o
D
November 1987 Am .J Obstet Gynecol
0
10
10
't: G
.,
=
•,.
u
•c
::J
6
6
E
"'I
N
'"'c
-
ii
3
3:
..
..
'-
G
Ill
...
Smokers
~ Non-Smokers
0
j:
280
14
Smokers Non-Smokers
240
c
N
u II a:
.
220
::J
"' 200 2
2
0
Smoker• Whole Blood Cd
Placental Cd
Placental Zn
Fig. 1. The effect of smoking on whole blood cadmium, placental cadmium, and placental zinc levels.
findings are in agreement with previous studies concerning cadmium exposure in pregnant women who smoke."· 10' 12 The zinc status of smokers and nonsmokers and that of their infants (as assessed by measurements of plasma and red blood cell zinc levels and alkaline phosphatase activity) is compared in Table IV. No significant differences were observed in the mothers. Moreover, diet histories verified that there was no difference in zinc intake between pregnant women who smoked and those who did not. The average daily intake of zinc was 10.4 mg. A major finding of the study was the 9% decrease in the red blood cell zinc level in the cord blood of infants of mothers who smoke (Fig. 2); this difference is statistically significant. A comparison of the zinc status of the mothers and infants shows that the infants had plasma zinc and alkaline phosphatase levels that were approximately one third higher than those of their mothers; on the other hand, the red blood cell zinc levels of the infants were approximately five times lower than those observed in the mothers. Comparisons of the levels of zinc in the placentas of smokers and nonsmokers are shown in Table IV and Fig. 1. Significantly higher levels of zinc were found in the placentas of pregnant women who smoked. The significantly higher levels of placental zinc correlated with significantly higher levels of alkaline phosphatase in the placenta. In smokers, the significant predictor of both placental zinc and placental cadmium by stepwise multiple regression was maternal whole blood cadmium. In nonsmokers, the maternal whole blood cadmium level pre-
NonSmoker•
Fig. 2. Cord vein red blood cell zinc levels in smokers and nonsmokers.
dieted the placental cadmium level but not the placental zinc level. These data are summarized in Table V, which shows the F value, degrees of freedom, and p value for each analysis. Since a 9% decrease in the red blood cell Zn level was observed in the cords of infants of mothers who smoke, the relationships between the biochemical variables and cord vein red blood cell zinc levels were examined. Significant predictors of the cord vein red blood cell zinc level, as assessed by stepwise multiple regression, were different in the smoking and nonsmoking groups. In smokers, thiocyanate was the strongest predictor of the cord vein red blood cell zinc level. The simple correlation showed that the relationship was negative, that is, the more thiocyanate in maternal plasma, the less zinc in the red blood cells of the infant. In nonsmokers, placental zinc was the strongest predictor of cord vein red blood cell zinc. The simple correlation showed that the relationship was positive, that is, the more placental zinc, the more cord vein red blood cell zinc. These data are summarized in Table VI, which shows the F value, degrees of freedom, p value, and r value for each analysis.
Comment The principal purpose of this study was to determine whether the increased levels of cadmium found in pregnant women who smoke affects the distribution of zinc in the maternal-fetal-placental unit. Our data show that increased cadmium levels as the result of smoking influence the levels of zinc in the placenta and cord blood and that this effect of cadmium is significant even though the absolute increase in cadmium as the result of smoking is small. To determine the effect of increased cadmium ex-
Volume 157 Number 5
Placental and fetal zinc status
1245
Table V. Significant relationships between maternal whole blood cadmium and placental cadmium or zinc in smokers and nonsmokers obtained with stepwise multiple regression Smokers Placental cadmium Placental zinc
F[2,59] F[l,60]
= =
Nonsmokers
4.6; p < 0.05 15.3; p < 0.001
F[l,83] = 4.7; p < 0.05 F[2,82] = 3.0; NS
Table VI. Predictors and univariate correlation coefficients of cord vein red blood cell zinc in smokers and nonsmokers Predictor Stepwise multiple regression Univariate correlation coefficient
Smokers
Nonsmokers
Thiocyanate F[2,58] = 4.14; p < 0.05 r = 0.21; p < 0.05
Placental zinc F[2,82] = 5.95; p < 0.025 r = 0.21; p < 0.05
posure on zinc status in pregnant women who smoke, we first verified that our study population of pregnant smokers and nonsmokers (with similar clinical characteristics) had similar levels of zinc intake. This was done by comparing diet histories and several indices of zinc status of the two groups. Our data showed that there was no difference in zinc intake between the smokers and the nonsmokers and that there was no difference in plasma zinc and red blood cell zinc levels. The data also showed that zinc intake of both groups was about 50% below the Recommended Daily Allowance for zinc during pregnancy, as reported by other investigators studying healthy pregnant patients. 5 However, since both groups had decreased zinc intake and were similar in all other respects, a comparison could be made in regard to the effect of cadmium exposure on zinc status. The cadmium exposure of our study population as the result of smoking was similar to that reported by us 6 and by others. 10"12 We observed a 64% increase in whole blood cadmium and a 45% increase in placental cadmium in the present study. The percent increase in placental cadmium reported to date has ranged from 26 10 to 400. 11. 12 The values differ, of course, depending on the study population and on the analytic techniques used. However, all the studies agree that there is a substantial increase in cadmium in the placentas of pregnant women who smoke. Increased exposure to cadmium and decreased zinc intake, as found in our pregnant smokers, are conditions that favor an effect of smoking on zinc status. Data supporting this are based on studies involving animals. For example, studies by Webb and Samarawickrama13 on rats show that maternal cadmium exposure causes fetal zinc deficiency and that this is one mechanism of the teratogenic effects of cadmium. By giving cadmium to pregnant rats these workers were able to show a dose-related decrease in fetal uptake of
zinc. Additional evidence that cadmium can induce fetal zinc deficiency was reported by Daston. 11 This author showed that coadministration of zinc almost eliminated the severe lesions in the fetal lung when the pregnant rat was given cadmium. Although the levels of cadmium in these studies are much higher than we observed in our study population, the threshold for such effects has not been established for animals or humans. Moreover, our data suggest that a cadmium/zinc interaction takes place in the maternal-fetalplacental unit of pregnant women who smoke. One aspect of our data that suggests a cadmium/zinc interaction and an effect of smoking on zinc status is the 10% increase in the level of zinc in the placentas of pregnant women who smoke. This is suggestive because it is likely that the increased levels of cadmium as the result of smoking induce the production of metallothionein, a low molecular weight protein that binds both cadmium and zinc. With the increased production of metallothionein there is an increased binding of cadmium and zinc and consequently a sequestering of these metals in the placenta. Although this has not been specifically studied in the placenta, autopsy studies of the kidney have shown that zinc levels increase in the kidney cortex with increasing concentrations of cadmium, and this increase is almost equimolar to the increase in cadmium. 15 Support for this interpretation is also provided by our statistical analysis, which shows that in smokers maternal whole blood cadmium levels are predictive not only of the placental cadmium levels but also of the placental zinc levels. Another aspect of our data that suggests an effect of smoking on zinc status is the 9% decrease in red blood cell zinc observed in the cords of infants of mothers who smoke. This decrease was found to correlate with the levels of thiocyanate (an index of the number of cigarettes smoked) in maternal blood. Thus the greater the number of cigarettes smoked, the lower the level
1246
Kuhnert et al.
of red blood cell zinc observed in cord blood. An explanation for this, which is supported by our data, is that zinc bound to metallothionein is accumulating in the placenta and that this binding may reduce the amount of zinc available to the fetus. Another possibility is that less zinc is being transported by the placenta to the fetus; this has been shown to occur in cadmiuminjected pregnant rats. 16 Regardless of the explanation, however, the data show that a cadmium/zinc interaction occurring in pregnant women who smoke may be harmful to the fetus since the decrease in cord red blood cell zinc is related to decreased birth weight in the infants of smokers. 17 Finally, additional evidence of the effect of smoking on fetal zinc status was noted in our data when the relationship between placental zinc levels and cord vein red blood cell zinc levels was examined. The data show a positive correlation between placental zinc and cord vein red blood cell zinc in nonsmokers; in smokers, however, this relationship was not found. The cadmium/zinc interaction, that is, the altered metabolism of zinc brought about by increased cadmium intake due to cigarette smoking, may explain the lack of this correlation in smokers. The effect of smoking on zinc status in the placenta and fetus may have no health significance if the availability of zinc is above the real requirement. On the other hand, this interaction may be detrimental to the fetus where zinc status is marginal. Additional studies are needed to clarify this point and to determine if zinc supplementation can counteract these changes in tissue levels (as it does in animals). Nevertheless, our studies show that there is an adverse effect of smoking on zinc status in the placenta and fetus, that is, a cadmium/zinc interaction does take place in the maternal-fetalplacental unit of pregnant women who smoke. This finding provides one additional reason for discouraging pregnant women from smoking. REFERENCES I. Lindskog S, Malmstrom BG. Metal binding and catalytic
activity in bovine carbonic anhydrase. J Bioi Chern 1962;237:1129.
November 1987 Am .J Obstet Gynecol
2. Plocke DJ, Levinthal C, Vallee BL. Alkaline phosphatase of Escherichia coli: a zinc metalloenzyme. Biochemistry 1962; 1:373. 3. Vallee BL. Biochemistry, physiology, and pathology of zinc. Physiol Rev 1959;39:443. 4. Parzyck DC, Shaw SM, Kessler MV, Vetter RJ, van Sickle DC, Mayes RA. Fetal effects of cadmium in pregnant rats on normal and zinc-deficient diets. Bull Environ Contam Toxicol 1978; 19:206. 5. Apgar]. Zinc and reproduction. Ann Rev Nutr 1985; 5:43. 6. Kuhnert PM, Kuhnert BR, Bottoms SF, Erhard P. Cadmium levels in maternal blood, fetal cord blood, and placental tissues of pregnant women who smoke. AM J OBSTET GYNECOL 1982;142:1021. 7. Kuhnert PM, Erhard P, Kuhnert BR. Analysis of cadmium in whole blood and placental tissues by furnace atomic absorption spectroscopy. In: Hemphill DD, ed. Trace substances in environmental health. Columbia, Missouri: University of Missouri Press, 1982 vol 16:370. 8. Pettigrew AR, Fell GS. Simplified colorimetric determination of thiocyanate in biological fluids, and its application to investigation of the toxic amblyopias. Clin Chern 1972; 18:996. 9. Doellgast GJ, Meis PJ. Use of specific inhibitors to discriminate alkaline phosphatase isoenzymes originating from human liver, placenta and intestine: absence of meconial alkaline phosphatase in maternal serum. Clin Chern Acta 1979;25: 1230. 10. Roels H, Hubermont G, BuchetJP, Lauwerys R. Placental transfer of lead, mercury, cadmium, and carbon monoxide in women. Environ Res 1978; 16:236. 11. Miller RK, Gardner KA. Cadmium in the human placenta: relationship to smoking. Teratology 1981 ;23:51. 12. Miller RK, Kellogg CK. The pharmacodynamics of prenatal chemical exposure. Bethesda, Maryland: National Institute on Drug Abuse, 1985:39. (Research monographs; series 60). 13. Webb M, Samarawickrama GP. Placental transport and embryonic utilization of essential metabolites in the rat at the teratogenic dose of cadmium. J Appl Toxicol 1981; 1,5:270. 14. Daston GP. Fetal zinc deficiency as a mechanism for cadmium-induced toxicity to the developing rat lung and pulmonary surfactant. Toxicology 1982;24:55. 15. Nordberg M, Elinder CG, Rahnster B. Cadmium, zinc, and copper in horse kidney metallothionein. Environ Res 1979;20:341. 16. Samarawickrama GP, Webb M. In: Proceedings of international conference on cadmium. Bethesda, Maryland: National Institutes of Health, 1978. 17. Kuhnert BR, Kuhnert PM, Debanne S, Williams TG. The relationship between cadmium, zinc, and birth weight in pregnant women who smoke. AM J 0BSTET GYNECOL 1987;157:1247.