Content uploaded by Rodney C Go
Author content
All content in this area was uploaded by Rodney C Go
Content may be subject to copyright.
678 Am J C/in Nutr 1990;5 1:678-84. Printed in USA. © 1990 American Society for Clinical Nutrition
A positive association between maternal serum zinc
concentration and birth weight13
Yasmin H Neggers, Gary R Cutter, Ronald TActon, Jose 0 Alvarez, Judith L Bonner,
Robert L Goldenberg, Rodney CP Go, and Jeffrey M Roseman
ABSTRACT A study was conducted on a cohort of 476
women (364 black, 1 12 white), who attended the Jefferson
County Health Department clinic for their prenatal care, to as-
certain the relationship between maternal serum zinc concen-
tration measured early in pregnancy and birth weight. For all
subjects maternal serum zinc was significantly related to birth
weight after various independent determinants of birth weight
were controlled for. The data in this study indicate a threshold
for maternal serum zinc concentration below which the preva-
bence of low birth weight increases significantly. Pregnant
women who had serum zinc concentrations in the lowest quar-
tile had significantly higher prevalence ofbow birth weight than
did those mothers who had serum zinc concentrations in the
upper three quartiles during pregnancy. These findings suggest
that maternal serum zinc concentration measured early in
pregnancy could be used to identify those women at higher risk
ofgiving birth to a low-birth-weight infant. Am J C/in Nutr
1990;S 1:678-84.
KEY WORDS Birth weight, low birth weight, serum zinc,
pregnancy
Introduction
Zinc is a trace element necessary for normal growth and de-
velopment. It is essential for cellular growth, division, and
differentiation (1). Requirements for zinc increase during peri-
ods of rapid growth, such as pregnancy, infancy, and puberty
(2). Animal studies showed that-dietary deficiency ofzinc dur-
ing pregnancy is associated with specific abnormalities and
growth retardation of the fetus (3, 4). In rats, maternal zinc
deficiency led to young that weighed 50% less than the con-
trols (5).
Recent evidence from human populations suggests the pro-
portion of malformations and other poor pregnancy outcomes
may be higher in populations where zinc deficiency has been
recognized (6). Ifthere is a relationship between maternal zinc
nutriture and birth weight, such information would be useful
because birth weight is an important factor that affects neonatal
mortality and is a significant determinant of infant and child-
hood morbidity (7).
Studies regarding the relationship between birth weight and
serum zinc concentration yielded conflicting results; some
were positive (8-1 1), some were negative (12, 13), and some
showed no association (14-1 7). Among the possible explana-
tions for the inconsistency in the results of the studies are 1)
zinc may have its effect primarily at only one time during preg-
nancy, 2) the naturally occurring physiologic decline in serum
zinc concentration during pregnancy might obscure the rela-
tionship when women are sampled at different times in preg-
nancy, and 3) there is a threshold for the effect of serum zinc
concentration. Most ofthe investigations were carried out with
serum zinc measured either during mid or late pregnancy. It
was suggested that zinc deficiency has the most profound effect
on rapidly proliferating tissue such as the embryo (18), hence
zinc nutriture may be ofgreatest importance during early preg-
nancy and studies that sample women later in pregnancy may
miss the association. During midpregnancy maternal zinc con-
centration is influenced by the heterogeneity regarding the tim-
ing and extent of hemodilution, which significantly lowers se-
rum zinc measures (19). This might mask or reverse the associ-
ation between maternal zinc status and birth weight. Finally, it
is possible that there may be a threshold relationship between
maternal serum zinc concentration and birth weight and that
the effect is only seen in populations ofmarginal zinc nutriture.
The measurement of zinc status among women of low socio-
economic status (who are more likely to have marginal pre-
pregnancy zinc nutriture) may detect such a threshold.
To address these issues, we measured serum zinc concentra-
tion relatively early in pregnancy in a large sample of women
ofbow socioeconomic status.
Subjects and methods
Subjects
The primary hypothesis ofthis study is that maternal serum
zinc concentration early in pregnancy is associated with birth
weight. This hypothesis was tested by means of a retrospective
IFrom the Departments ofHuman Nutrition and Hospitality Man-
agement, The University ofAbabama, Tuscaloosa, AL, and the School
of Public Health, the Department ofObstetrics, and Gynecology, and
the Department of Microbiology, University of Alabama at Birming-
ham, Birmingham, AL.
2Supported in part by grant DK 32767 from the National Institutes
ofDiabetes and Digestive and Kidney Diseases.
3Address reprint requests to YH Neggers, P0 Box 870158, Univer-
sity ofAlabama, Tuscaloosa, AL 35487-0158.
Received December 18, 1988.
Accepted for publication June 7, 1989.
by guest on July 11, 2011www.ajcn.orgDownloaded from
SERUM ZINC AND BIRTH WEIGHT 679
41± SD (range).
t Significantly different, p = 0.02.
TABLE 1
Description ofpregnant women enrolled in the study
All subjects White Black
n x± SD (range) ni± SD (range) n ± SD (range)
Age(y) 476 21.8± 4.9(12-42) 112 21.0± 4.5(14-36) 364 22.1 ± 5.0(12-42)
Prepregnancyweight(kg) 341 61.7± 13.8(30-120) 90 60.8± 14.0(37-120) 251 62.1 ±13.7(30-1 10)
Gestational age at first
visit(wk) 476 16.0± 5.2(6-31) 112 15.1± 4.8(8-27) 364 16.3± 5.3(6-31)
Smoking
Yes 117(25.2%) - 55(49.5%) -61(17.2%) -
No 348 (74.8%) -56 (50.5%) -293 (82.8%) -
Alcohol consumption
Yes 56(12%) - 13(11.7%) - 43(12.1%) -
No 409 (88%) -98 (88.3%) -3 11 (87.9%) -
cohort study design. The sample for the study consisted of preg-
nant women (476) who had blood drawn twice during the preg-
nancy while they received prenatal care at the Jefferson County
Health Department for tht 6-mo period from June to Novem-
ber 1984. The protocol for the study was approved by the Insti-
tutionab Review Board for Human Use ofthe University of Al-
abama at Birmingham.
Serum anali’ses and other data
Maternal serum zinc concentration was the exposure vari-
abbe. A 5mL blood sample was collected in an evacuated tube
from each subject during her visit to the Health Department
and analyzed for zinc concentration by a standard atomic-ab-
sorption spectrophotometric procedure (20) with an atomic-
absorption spectrophotometer (model 372, Perkin-Elmer Cor-
poration, Norwalk, CT). The outcome variable measured was
birth weight. The mean serum zinc concentrations in this study
are in the range reported by others(2 1). The coefficient of varia-
tion obtained by replication of 6 aliquots was 4.2%. The con-
founding or interacting variables examined were estimated ges-
tational age at birth, maternal age, prepregnancy weight, weight
gain during pregnancy, race, smoking, and alcohol consump-
tion. For this study smoking and alcohol consumption were
considered positive if a woman indicated that she was using
either substance in any quantity at the time ofthe first prenatal
visit. Information on birth weight and the various confounding
factors was obtained from the OBAR system (22), which is a
computerized obstetrical record system designed to follow
health department women through their pregnancies and dcliv-
eries at either University ofAbabama at Birmingham or Cooper
Green Hospital, the official county hospital.
Statistical analyses
Mean birth weight, gestational age at birth, maternal serum
zinc concentration, and other maternal characteristics were es-
timated separately for blacks and whites and were compared
with the Student’s ttest. Pearson correlation coefficients were
computed between serum zinc concentration and birth weight
and various maternal characteristics. Subjects were divided
into quartiles of serum zinc concentration (adjusted for esti-
mated gestationab age at the time ofblood draw) for categorical
analyses. Newborns were classified into two groups, low birth
weight (< 2500 g) and normal birth weight ( 2500 g). The
measure of association computed was the prevalence odds ra-
tio. To assess the contribution ofserum zinc to explain the vari-
ation in birth weight or low birth weight after the known con-
founding factors noted above were controlled for, multivari-
able linear and logistic regression analyses were performed
(SAS, SAS Institute, Cary, NC).
Results
Table I gives the description of pregnant subjects. There
were 76.4% blacks and 23.6% whites. The differences in age and
prepregnancy weight between blacks and whites were small and
not significant. The mean gestationab age at the first visit for
whites was significantly earlier(p = 0.05) than for blacks. There
was considerable difference in smoking (p = 0.0001) yet little
difference in alcohol consumption (p = 0.9 1) between whites
and blacks.
Serum zinc
The mean serum zinc concentration of all subjects is given
in Table 2. The mean serum zinc concentration for whites was
significantly higher than that for blacks. The mean serum zinc
concentrations by week of gestation are given in Figure 1 for
all subjects. In Table 3 serum zinc values are grouped by time
ofbbood draw. Zinc values appeared to be bower in women who
were seen for the first time in later stages ofgestation. To evalu-
ate the effect ofweek ofgestation on serum zinc concentration,
a regression was performed on the serum zinc concentration
vs the gestational age at the time of blood draw. Serum zinc
concentration was significantly rebated to the gestational age at
TABLE 2
Serum zinc concentrations ofall subjects and by race
nZinc concentration4
imo//L
Allsubjects 476 14.1 ±2.8 (6.4-25.7)
Whites 1 12 14.7 ±3.Ot (6.4-22.9)
Blacks 364 13.9 ±2.8t (6.4-25.7)
by guest on July 11, 2011www.ajcn.orgDownloaded from
4±SD.
680
20
18
16
0
E14
C,
.-N
E
12
‘S
a,
10
8
6
NEGGERS ET AL
4 6 8 10 12 14 16 18 20 22 24 26 26 30 32 34 36
Weeks of gestation
FIG I.Mean serum zinc concentration by weeks ofgestation for all subjects (3-wk moving average).
the time ofthe blood draw (j = -0.09, p= 0.0002, R2 = 0.03). which blood was collected. Note that there were no important
Serum zinc concentration was adjusted by time of the blood differences between the results ofanalyses when unadjusted se-
draw to eliminate the effect of differences because of time at rum zinc values were used. Including the square of gestational
TABLE 3
Serum zinc concentrations grouped by weeks of gestation
Weeks of
gestation
All subjects Whites Blacks
nZinc concentration4 nZinc concentration nZinc concentration
Mmo//L Mmo//L zmo//L
6-9 47 14.0±3.1 17 14.6±3.0 30 13.7±3.1
10-14 166 14.6±2.9 39 14.6±2.7 127 14.6±2.9
15-19 146 14.1 ±2.7 35 15.7±3.1 111 13.6±2.4
20-24 87 13.5 ± 2.7 16 12.8 ± 2.6 71 13.6 ± 2.7
25-29 25 13.3 ±2.6 5 14.1 ±3.1 20 13.1 ±2.5
30-31 511.4±1.7 - - 5 11.4±1.7
by guest on July 11, 2011www.ajcn.orgDownloaded from
Serum zinc (umol/L)
SERUM ZINC AND BIRTH WEIGHT 681
4Partial regression coefficient.
FIG 2. The distribution ofserum zinc concentrations by quartiles for all subjects.
age at the time ofblood draw did not significantly improve the
model. In all other regressions involving serum zinc concentra-
tion, this adjusted value was used. Figure 2 presents the distri-
butions ofadjusted serum zinc values by quartiles for all sub-
jects.
Regression analyses of serum zinc concentration on mater-
nal age, prepregnancy weight, weight gain during pregnancy,
alcohol consumption, and smoking showed no significant rela-
tionships (Table 4). On the other hand, serum zinc concentra-
tion was found to be significantly related to gestational age at
birth by linear regression (/3 = 0. 1 7, p= 0.0001, R2 = 0.05).
The relationship between serum zinc concentration and vari-
ous maternal characteristics was also evaluated by race. Results
were similar to that for all subjects, ie, for both blacks and
whites, serum zinc concentration was significantly related only
to gestationab age at birth.
Birth i’eig/zt
Table 5shows the distribution ofbirth weight and gestational
age at birth for all subjects and by race. The mean birth weight
of white infants was significantly higher than that of black in-
fants (p -0.03) even after gestational age at birth was con-
trolled for. Of the births, 8.2% were low birth weight (< 2500
g) for all subjects. Among whites 9.2% ofinfants were low birth
weight whereas 7.9% of black infants were low birth weight.
TABLE 4
Relationships between various maternal characteristics and serum
zinc concentrations
Variable * SEE p
Age -0.02 0.03 0.53
Weight gain during pregnancy -0.2 1 0. 150.1S
Prepregnancy weight 0.01 0.01 0.38
Alcohol -0.24 0.40 0.54
Smoking 0.29 0.29 0.32
by guest on July 11, 2011www.ajcn.orgDownloaded from
682 NEGGERS ET AL
4Final regression model for prediction ofbirth weight by using stepwise regression.
1 Partial regression coefficient.
TABLES
Characteristics of newborns
All subjects Whites Blacks
ni±SD(range) ni±SD(range) ni±SD(range)
Birth weight (g) 462 3188.3 ±561.2 (420-5280) 109 3335.3 ±635.6 (420-4280) 353 3142.9 ±529.1 (500-5280)
Gestational age (wk) 476 39.2 ±2. 1 (22-44) 1 12 39.7 ±2.3 (26-44) 364 39. 1 ± 2.0(22-44)
There was no significant difference in proportion of low-birth-
weight infants between blacks and whites (p = 0.68).
Birth weight and serum zinc concentration
There was a significant correlation between birth weight and
serum zinc concentration for all subjects (y = 0.32, p
= 0.0001), whites (-y = 0.38, p-0.001), and blacks (-y 0.28,
p= 0.000 1 ). To assess whether serum zinc concentration was
associated with birth weight because of confounding that was
due to sex ofthe infant or to other known maternal charactens-
tics, a stepwise multiple regression employing the sex of the
infant, maternal age, race, prepregnancy weight, weight gain
during pregnancy, gestational age at birth, smoking, and alco-
hol consumption was used. For all subjects and for both whites
and blacks, serum zinc concentration was significantly related
to birth weight after various independent determinants of birth
weight were controlled for (Table 6). The results of the final
regression model, with those independent determinants of
birth weight that were significant at p0.05 in the multiple-
regression model, are shown in Table 6 for all participants and
for whites and blacks.
Low birth weight and serum zinc concentration
To assess the relationship between low serum zinc concen-
tration and low birth weight, multiple logistic regression was
used to calculate prevalence odds ratios for low birth weight
between various quartiles of serum zinc concentration; the
highest quartile of serum zinc concentration was used as the
reference category while gestational age at birth and other mdc-
pendent determinants of birth weight were controlled for (Ta-
ble 7). The prevalence of low birth weight was eight times
higher among women with serum zinc concentration in the
TABLE 6
lowest quartile than for women with serum zinc in the highest
quartile. On the other hand, the prevalence odds ratio was 5.8
(3. 1 , 1 3.5) when women with serum zinc concentration in the
lowest quartile were compared with women with serum zinc
concentration in all other quartiles combined.
Discussion
The relationship between serum zinc concentration and
birth weight was evaluated in a large biracial sample of lower-
socioeconomic-status women. A significant independent, posi-
tive association was observed between serum zinc concentra-
tions and birth weight overall and separately for both races. In
particular, the serum zinc concentrations in the lowest quartile
were associated with eight times the frequency of low birth
weight as compared with the frequency in the highest quartile.
Findings vary regarding the relationship between maternal
serum zinc concentration and gestational age at birth. Jameson
(8) and Kiibholma et al (23) reported a positive association be-
tween maternal serum zinc concentration and gestationab age
at birth. McMichael et al (12) reported a weak and statistically
nonsignificant inverse association whereas Cherry et al (16)
found no association between gestational age at birth and ma-
ternab serum zinc concentration. In the present study, serum
zinc concentration was significantly positively associated with
gestationab age at birth. Because gestational age is a known pre-
dictor ofbirth weight (24), to control for this effect the relation-
ship between serum zinc concentration and birth weight was
studied after the effect of gestational age on birth weight in the
multiple-regression model was controlled for. Serum zinc con-
centration was rebated to gestational age at the time when zinc
Relationships between maternal serum zinc concentrati ons and bi rth weight4
All subjects Whites Blacks
Variable
(R2 0.27: n= 327) (R2 = 0.30; n=87) (R2 = 0.24: n= 240)
Coefficientt SEE p Coefficient SEE p Coefficient SEE p
Intercept -668.6 - - -230.4 -- -113.0 - -
Gestationalage 85.1 13.1 0.0001 85.8 30.8 0.006 86.3 14.2 0.0001
Race -232.3 62.6 0.0002 - - ----
Prepregnancyweight 5.6 1.9 0.002 - - - 7.3 2.2 0.001
Smoking -176.0 61.9 0.005 -273.5 104.7 0.01 ---
Sexofinfant -1 16.4 52.2 0.03 -297.9 106.1 0.006 - - -
Alcohol - - -- - --172.4 84.4 0.04
Adjustedzinc 37.9 9.1 0.0001 52.9 18.3 0.005 28.7 10.4 0.008
by guest on July 11, 2011www.ajcn.orgDownloaded from
SERUM ZINC AND BIRTH WEIGHT 683
weight and serum zinc concentration measured later in preg-
TABLE 7
Prevalence odds ratios by serum zinc concentration quartiles for all
subjects(n =327)
Quartile t* SEE x2 t PORT 95%ClI
Lowest(7.0-12.2Mmol/L) 2.1 0.62 I 1.1 0.001 8.2 2.4, 27.5
Second ( I 2.2- 1 3.7 Mmol/L) 0.62 0.68 0.83 0.36 I .8 0.49, 7.0
Third(13.7-15.9MmoI/L) 0.24 0.75 0.10 0.74 1.2 0.29,5.5
Highest(15.9-25.4zmol/L) -- - -1.0 -
aUnconditional maximum likelihood estimate of coefficient.
t Two-sided p value.
tPrevalence odds ratio (after controlling for gestational age, race, and other
predictors ofbirth weight).
§Confidence interval.
was measured. Therefore, serum zinc concentration was ad-
justed for the time ofthe blood draw.
The decrease in serum zinc concentration with increase in
gestational age at blood draw has two possible interpretations.
The first is plasma volume expansion and hypoalbuminemia,
which have been cited frequently (14, 19, 25, 26) as possible
causes of the fall in plasma zinc with the progression of preg-
nancy. The other has to do with the observation that women
who seek prenatal care later in pregnancy have a greater fre-
quency ofbow-birth-weight babies (27). It is possible then, that
ifthe same women had been studied earlier during pregnancy,
they would still have had comparatively low serum zinc con-
centrations. Thus by adjusting the serum zinc concentration by
week of gestation, as has been done in this study, the serum
zinc concentrations of those women who provided the blood
sample later during gestation was raised artificially. This should
bead to a bias toward the null. Nevertheless, the data in this
study indicate that the relationship between birth weight and
maternal serum zinc concentration remains unchanged even
after this adjustment.
Previous epidemiobogicab studies regarding the relationship
between maternal serum zinc concentration and birth weight
are contradictory and inconclusive. Researchers indicated pos-
itive (8- 1 1), negative (1 2. 13), and no correlation (14- 17) be-
tween maternal zinc nutriture and birth weight. Why is there a
strong relationship in only a few studies? The design and condi-
tions of different studies were different and sample size was
small in some cases. Maternal serum zinc concentration varies
with gestational age (25). In most studies zinc was measured
either during midpregnancy or late in pregnancy. There is a
large increase in plasma volume during the second trimester
and a plateau is reached at --34 wk ofgestation (28). Therefore,
around midpregnancy considerable hemodilution will have oc-
curred. Because there is a significant variability in the rate of
increase of blood volume, at any midpregnancy measuring
point these differential increases in plasma volume may cause
the plasma zinc values to vary considerably from subject to
subject. This could mask the relationship between serum zinc
concentrations and birth weight or even reverse the direction
of the relationship. A smaller plasma volume increase was re-
ported in women with intrauterine fetal growth retardation
than in those with normal pregnancies (29). This factor could
also explain the inverse correlation observed in some studies
between plasma zinc concentration and birth weight (12, 13).
Another explanation of inverse relationship between birth
nancy is the increased fetal accretion of zinc by larger infants
as they grow.
The results ofa study by Jameson (8) where serum zinc con-
centration was measured early in pregnancy agree with the
present study. The mean gestational age when zinc was deter-
mined (14 wk) in Jameson’s study is close to our samples (16
wk). Jameson also reported a positive relationship between the
serum zinc concentration and birth weight and the serum zinc
concentration and the length ofgestation; birth weight was not
adjusted for gestationab age.
The interpretation that maternal serum zinc concentration
is a predictor of birth weight is further supported by the fact
that the data in this study indicate that there is a threshold for
maternal serum zinc concentration below which the preva-
lence of low birth weight increases considerably. Pregnant
women who had serum zinc concentrations in the lowest quar-
tile had significantly higher prevalence ofbow birth weight than
did women who had serum zinc concentrations in the upper
three quartiles during pregnancy (Table 7). Conversely, women
with serum zinc concentration in the second and third quartiles
did not show any significant increased risk of low birth weight
as compared with women with serum zinc concentration in the
highest quartile.
This large study shows a significant positive association be-
tween maternal serum zinc concentration and birth weight in-
dependent of other known risk factors. It is possible that the
association reported is seen only in women of lower socioeco-
nomic status who have marginal zinc nutriture during early
pregnancy. It is also possible that the association between se-
rum zinc concentration and birth weight is secondary to an-
other causal factor. One possibility is albumin concentration.
Zinc in serum is bound primarily to albumin (19), and hypoal-
buminemia may possibly be related to growth retardation; both
should be measured to resolve this question.
The results ofthis study are striking. Low serum zinc concen-
tration was a more significant predictor of low birth weight
than almost all known risk factors for low birth weight. If the
results are replicable, they will have important implications for
both the prediction of low birth weight and, possibly, its pre-
vention. B
References
1. Elizaga V. Ferreira RMCD. Zinc, pregnancy and parturition. Acta
Pediatr Scand [Supplj 198:319:150-7.
2. Jameson S. Zinc and pregnancy. In: Nriagu JO, ed. Zinc in the
environment. Part II: Health effects. New York: John Wiley and
Sons, 198 1:183-97.
3. Hurley LS. Teratogenic aspects of mangnese, zinc and copper nu-
trition. Physiol Rev 1981:61:249-95.
4. Hickory W, Nanda R, Catalanotto FA. Fetal skeletal malforma-
tions associated with moderate zinc deficiency during pregnancy.
JNutr 1979; 109:883-91.
5. Hurley LS, Swenerton H. Cogenital malformations resulting from
zinc deficiency in rats. Proc Soc Exp Biol Med 1966: 123:692-7.
6. Worthington-Roberts B. The robe ofnutrition in pregnancy course
and outcome. J Environ Pathol Toxicol Oncol 1985:5(6): 1-80.
7. McCormick MC. The contribution of low birth weight to infant
mortality and childhood morbidity. N Engb J Med 1985:312:82-
90.
8. Jameson S. Zinc and copper in pregnancy. Correlation to fetal and
maternal complications. Acta Med Scand [Suppl] 1976: 593:5-20.
by guest on July 11, 2011www.ajcn.orgDownloaded from
684 NEGGERS ET AL
9. Crosby WM, MetcoffJ, Costiloe JP, et al. Fetal malnutrition. An
appraisal ofcorrelated factors. Am J Obstet Gynecol 1977; 128:22-
31.
10. Mameesh MS. Hathout H, Safar MAAI, Mahfouz A. Maternal
plasma proteins, magnesium, zinc and copper concentrations at
term associated with birth size in Kuwait. Ada Vitaminol Enzy-
mol 1985:7(3-4): 183-8.
1 1. Atinmo T, Mbofung C, Osinusi BO. Relationship ofzinc and cop-
per concentrations in maternal and cord blood and birth weight.
Int J Gynecol Obstet 1980; 18:452-4.
12. McMichael AJ, Dreosti bE, Gibson GT, Hartshore JM, Buckley
RA, Colley DP. A prospective study ofserial maternal serum zinc
levels and pregnancy outcome. Early Hum Dev l982;7:56-69.
13. Metcoff J, Castiloe JP, Crosby W, et al. Maternal nutrition and
fetal outcome. Am J Clin Nutr I 98 1;34:708-21.
14. Tuttle 5, Aggett PJ, Campbell D, MacGellivary I. Zinc and copper
nutrition in human pregnancy: a longitudinal study in normal pri-
migravidae and in primigravidai at risk ofdelivering a growth re-
tarded baby. Am J Clin Nutr 1985:41:1032-41.
15. Hunt FH, Murphy NH, Cleaver AE, et al. Zinc supplementation
during pregnancy: effects on selected blood constituents and on
progress and outcome ofpregnancy in low-income women of Mex-
ican descent. Am J Clin Nutr 1984:40:508-21.
16. Cherry FF, Bennett EA, Bazzano GS, Johnson DK, Fosmire GJ,
Batson HK. Plasma zinc in hypertension/toxemia and other repro-
ductive variables in adolescent pregnancy. Am J Clin Nutr
198 1:34:2367-75.
17. Prema K. Predictive value ofserum copper and zinc in normal and
abnormal pregnancy. Indian J Med Res 1980:71:554-60.
18. Dreosti lE. Zinc in prenatal development. In: Prasad AS, Dreosti
IA, Hetzel BS, eds. Clinical applications ofrecent advances in zinc
metabolism. New York: Alan R Liss, Inc. 1982:19-27.
19. Swanson CA, King JC. Reduced serum zinc concentration during
pregnancy. Obstet Gynecol 1983:62:313-8.
20. Perkin-Elmer Corporation. Analytical methods for atomic absorp-
tion spectrophotometry. Norwalk, CT: Perkin-Elmer Corporation,
1982.
2 1.Hambidge KM. Mauer AM. Trace elements. In: Committee on
Nutrition of Mother and Preschool Child. Laboratory indices of
nutritional status in pregnancy. Washington DC: National Acad-
emy ofSciences, 1978:167-74.
22. Wirtshafter DD, Blackwell WC, Goldenberg RL, et al. A country-
wide obstetrical automated medical record system. J Med Syst
1982;6:277-90.
23. Kiilholma P. Gronroos M, Erkkola R, Parkarinen P. Nanto V. The
robe ofcalcium, copper, iron and zinc in preterm delivery and pre-
mature rupture of fetal membranes. Gynecol Obstet Invest
1984; 17:194-201.
24. Koops BL, Morgan LI, Battaglia FC. Neonatal mortality risk in
relation to birth weight and gestational age. Update J Pediatr
1982: 10 1:969-77.
25. Hambidge KM. Krebs NF, Jacobs MA, Favier A, Guyette L, IkIe
DN. Zinc nutritional status during pregnancy: a longitudinal
study. Am J Clin Nutr 1983; 37:429-42.
26. Campbell-Brown PC, Ward Ri, Haines AP, North WRS, McFad-
yen IR. Zinc and copper in Asian pregnancies, is there evidence
for a nutritional deficiency? Am J Obstet Gynecol l985;92:875-
85.
27. Gartmaker SL. The effects ofprenatal care upon the health of the
newborn. Am J Public Health 1979:69:653-60.
28. Hutchins CJ. Plasma volume changes in pregnancy in Indian and
European primigravidae. BrJ Obstet Gynaecol l980;87:586-9.
29. Gibson HM. Plasma volume and glomeular filteration rate in preg-
nancy and their relation to differences in fetal growth. J Obstet
Gynaecol Br Commonw l973;80: 1067-74.
by guest on July 11, 2011www.ajcn.orgDownloaded from