Resting metabolic rate and body composition of healthy

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

Swedish women during pregnancy13
Elisabet Forsum, PhD; Aija Sadurskis, BSc; andfan Wager, MD

ABSTRACT Body weight, resting metabolic rate (RMR), total body potassium (TBK),
and total water were measured and total body fat (TBF) was calculated
body in a longitudinal
study of 22 pregnant, healthy Swedish women. Measurements were made before pregnancy,
at gestational weeks 16-18, 30, and 36, and 5-10 d and 6 mo postpartum. RMR increased
more during pregnancy than previous estimates on well-nourished women showed and the
increase was significantly correlated with the birth weight ofthe baby. TBK decreased during
the first part ofpregnancy; measurements at weeks 16-18 and 30 were significantly lower than
the prepregnancy value and changes in TBK and RMR were significantly correlated. TBF gain
during pregnancy was 5.8 ± 4.0 kg and 60% was already gained by gestational weeks 16-18.
Gain in fat was not correlated with birth weight. Am J Clin Nutr l988;47:942-7.

KEY WORDS Birth weight, energy requirements, pregnancy, reproduction, resting meta-
bolic rate, total body fat, total body potassium, total body water

Introduction quency ofperinatal deaths. Powerful mechanisms appar-

ently exist that, for the benefit ofreproduction, conserve
Our understanding of energy requirements during energy during pregnancy. This was demonstrated by
normal human pregnancy is far from complete. Hytten Lawrence et al (3), who measured RMR of unsupple-
and Leitch (1) estimated that the average well-nourished mented and supplemented malnourished Gambian
woman gaining 12.5 kg body wt during pregnancy women. They found that the cumulative energy costs of
requires 80 000 kcal in addition to her nonpregnancy increased maintenance metabolism were only 1000 kcal
energy needs. This figure includes 36 000 kcal for an for the unsupplemented and 13 000 kcal for the supple-
increased basal metabolic rate (BMR) to cover main- mented women. In a preliminary report (4) we showed
tenance energy requirements and 44 000 kcal for the that the corresponding figure for well-nourished Swedish
synthesis of new tissue including 3-4 kg body fat stored women is “-46 000 kcal, a figure considerably above ear-
in the maternal body. Although it seems likely that these her estimates for well-nourished populations of women.
figures are fairly realistic averages in healthy, well- It appears that the change in metabolic rate caused by
nourished women little is known about variations pregnancy varies considerably among populations with
among individuals. A successful outcome to pregnancy different nutritional status. Consequently, additional
can occur within a large range ofchanges in body weight data collected on well-nourished and malnourished
but little is known about the nature and quality of the women with respect to variables of interest for under-
materials stored. This was pointed out by Thomson (2),
who suggested that investigations should be directed at From the departments of Medical Nutrition and Obstetrics and
the nature and meaning of these variations. Further- Gynecology, Karolinska Institute, Huddinge University Hospital,
more, considerable uncertainty exists regarding the true Huddinge, Sweden.
changes in body weight, body composition, and resting 2 Supported by grants from the Swedish Medical Research Council
metabolic rate (RMR) during early pregnancy because (project #B86-l9x-0659 l-04A), the General Maternity Hospital Foun-
most studies have started at the 10th week of gestation dation, M Bergvall’s Foundation, Jeansson’s Foundation, Nestl#{233}
Foun-
dation, Swedish Nutrition Foundation, OE and E Johansson’s Founda-
or later.
tion, the Swedish Baby Food Industry Fund for Nutritional Research,
It appears that successful reproduction in malnour-
and F#{246}reningenMj#{246}lkdroppen.
ished women can occur on dietary intakes much below 3 Address reprint requests to Elisabet Forsum, PhD, Karolinska
those suggested by the data of Hytten and Leitch (1) al- Institute, Department of Medical Nutrition, Huddinge University
though such situations clearly are suboptimal for several Hospital F69, 5-141 86 Huddinge, Sweden.
reasons, including low average birth weight, increased Received February 25, 1987.
frequency oflow-birth-weight infants, and increased fre- Accepted for publication June 23, 1987.

942 Am J C/in Nuir l988;47:942-7. Printed in USA. © 1988 American Society for Clinical Nutrition
 ENERGY REQUIREMENTS DURING PREGNANCY 943

TABLE 1 Protocol
Characteristics ofstudy participants
Measurements were made of each woman at six times: be-
Prepregnancy body weight (kg) 6 1.0 ± 9.9 fore pregnancy, three times during pregnancy (l6th-l8th ges-
Height (cm) 165 ± 5 tational week, 30th gestational week, 36th gestational week),
Body mass index (kg/rn2) 22.3 ± 3.5 and 5-10 d and 6 mo postpartum. The gestational week was
Prepregnancy energy intake (kcal/d) 1900 ± 400 estimated from ultrasound measurements taken in week 16.
Measurements of body weight, body composition,

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

68 ± 15 and RMR
Prepregnancy protein intake (g/d)
Total weight gain during pregnancy (kg) I 3.6 ± 3.0
were carried out on each occasion. Food intake was measured
Length ofgestation (d) 275 ± 14 before pregnancy. At each measurement the woman arrived
Birth weight ofbaby (g) 3560 ± 443 by taxi at the hospital, =-07#{174}.She was instructed not to have
Age at prepregnancy measurement (y) 28.7 ± 4.0 breakfast (but had not necessarily been fasting for 12 h) and to
Time between prepregnancy measurement and start do as little as possible on the morning oftesting. First, she was
ofpregnancy (wk) 14.2 ± 9.3 given a dose of ‘80-labeled water to drink for estimating total
body water (TBW) and then she rested for 45-60 mm before
#{149}i±5Dn22 RMR and total body potassium (TBK) were estimated. The
protocol of this study was approved by the ethical committee
ofthe Karolinska Institute, Stockholm, Sweden.
standing nutritional requirements during pregnancy are
Mazernalbody weight
needed as a basis for establishing adequate dietary guide-
lines during pregnancy. Such studies may also help to Maternal body weight (MBW) ofwomen in light underwear
define an optimal level ofnutrition during reproduction was estimated with a Stathmos mechanical balance (AB
and to understand the energy-conserving mechanisms Stathmos, J#{246}nk#{246}ping,
Sweden) to the nearest 0. 1 kg. The same
apparently operating during pregnancy. balance was used throughout the study. The women were also
weighed in the delivery room before delivery and the difference
This study was designed to measure longitudinal
between this value and the prepregnant body weight was con-
changes in RMR and body composition in healthy Swe- sidered to be the total weight gained during pregnancy.
dish women starting before conception and ending 6 mo
after delivery. Totalbody water
Total body water (TBW) was estimated as described by
Subjects and methods Schoeller et al (6). The appropriate amount of ‘80-labeled wa-
ter(YEDA Stable Isotopes, Rehovot, Israel), 10 atomic percent
Enrollment of subjects (100 gIL) was weighed out. The subjects received 1 g/kg body
wt ofthis water at all measurements except the last pregnancy
In Sweden birth control is to a large extent provided at ma- measurement and the first measurement postpartum, when 0.4
ternity health
centers by midwives authorized to apply and re- g water/kg body wt was given. Baseline saliva and urine sam-
move intrauterine devices(IUDs)and prescribe oral contracep- ples were collected before the dose was given and after 4 h an-
tives. We arranged with midwives at Huddinge University Hos- other saliva sample was taken. Two to six morning urine sam-
pital’s maternity health centers to ask each woman wishing to ples were also collected from each subject during a period of
remove her IUD to become pregnant to consider participating -‘--2 wk after the dose was taken. 180 in saliva and urine was
in a study ofnutrition during pregnancy and lactation. If inter- analyzed with a mass spectrometer (Sira 9, VG Isogas, Man-
ested the woman was given details ofthe study and was invited chester, Great Britain) and body water was calculated from sa-
to participate. Seventy-seven women received this information liva samples collected before and 4 h after dosing. In 14 cases
and 33 agreed to participate. Four of them did not become these values were also checked by analyzing urine samples and
pregnant during the study and seven left the study for other extrapolating to time zero assuming single-pool kinetics (7).
reasons. Thus, 22 women completed the study. As soon as pos-
sible after enrollment the prepregnancy measurements were Total body potassium
made and the IUD was removed.
TBK was estimated in a whole-body counter (8) calibrated
Subjects with a phantom containing KG and corrections were made for
different weights and heights of the subjects. A counting time
Women were recruited from a middle-income population of 2000 5 was used for the patients and background, giving an
of Swedish citizens. All were married, had completed 9 y of overall methodological error of7-8% (9). TBK oftwo healthy
compulsory schooling and had some kind ofprofessional edu- subjects who had little variation in body weight were found to
cation. At the time of enrollment they were all professionally be within the expected range throughout the course of the
active (eg, in nursing, teaching, or clerical work). Four of the study.
women had two children, one had none, and the remaining 17
each had one child. A compilation ofanthropometnic, dietary,
Total bodyfat
obstetric, and general data about the 22 study participants is
shown in Table 1. They were all healthy, all had normal preg- Total body fat (TBF) was calculated with the model de-
nancies, and all but one delivered vaginally. All started breast- scribed by Pipe et al (10), which is based on estimates of MBW,
feeding in the hospital and 6 mo postpartum 12 ofthem were TBK, and TBW. During pregnancy this model deducts the
still partially breast-feeding. Some of the data obtained for 17 contributions of weight, water, and K of the fetus and its ad-
ofthese women were reported in a study ofenergy metabolism nexa, thus making the calculation ofTBF in the maternal body
during human lactation (5). possible.
944 FORSUM ET AL

TABLE 2
Maternal body weight (MBW), total body water(TBW), total body potassium (TBK), total body fat (TBF), and resting metabolic rate (RMR) in
healthy Swedish women before, during, and after pregnancy

MBW TBW TBK TBF RMR

kg % kg % mmol % kg % kcal/min

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

Prepregnancy 6 1.0 ± 9.9 33.0 ± 4.3 2397 ± 327 I 7.2 ± 6.9 0.936 ±0.124
Gestational
weeks
16-18 63.7± 97t 104.7±3.6 32.5±3.7 99.1 ± 8.5 2224±298 93.5± 11.6 20.7±6.Ot 128.3±28.7 1.000±0.157 107.8± 17.2
Gestational
week30 70.2± 9.9t 115.6±4.6 36.7±3.7t 1123± 11.6 2290±330 95.8± 9.0 22.6±6.9t 141.3±39.5 I.147±0.I43t 124.0± 18.1
Gestational
week36 72.7± 10.3t 119.8±5.1 38.7±4.4t 118.4± 13.3 2507±307 105.8± 14.4 22.3±7.lt 139.2±39.6 I.223±O.124t 132.5± 18.9
5-lOd post-
partum 67.6± 10.8t 111.1 ±5.7 34.2±3.71 104.3± 12.91 2366±294* 99.0± 1081 22.9±7.8t1 142.9±40.81 I.063±0.169* 114.1±23.81
6 mo post-
partum 61.9± 10.9k 101.7±5.80 31.4±4.0U 95.6± 7.31 2240±252 95.1 ± 8.3 20.4±7.91 123.2±28.91 O.984±O.103t1 108.6± 16.61

S SD. n - 22 except where otherwise indicated. % - percent ofvalue obtained before pregnancy.
t Significantly different from prepregnancy value (p < 0.001).
t Significantly different from prepregnancy value (0.05 > p > 0.01).
§n-21.
I Significantly different from prepregnancy value (0.01 > p > 0.001).
In = 20.

Resting metabolic rate showed a tendency to decrease during early pregnancy

RMR was estimated by collecting the expired air during two and was also significantly lower 6 mo after pregnancy
periods of 8-12 mm in Douglas bags (B#{246}rgeHansen, K#{246}pen- than before. Calculated figures for body fat show that
hamn, Sweden). Duplicate analyses ofoxygen and carbon diox- these women contained significantly more fat at all mea-
ide were made according to Scholander (1 1). Energy metabo- surements during and after pregnancy than before. By
lism was calculated according to Brouwer (12). The increase the first measurement during pregnancy the women had
over the prepregnant value in RMR was calculated for each gained an average of3.5 kg body fat (range, - 1 1 -9.8 kg), .

subject for each measurement and plotted against the gesta- constituting on average about 60% of the total fat gain
tional time. To calculate the cumulative increase in RMR, the
during pregnancy. On average these women gained 5.8
area under the curve from the start ofpregnancy until delivery
kg body fat during the complete pregnancy (range, -2.6-
was estimated.
12 kg) and 6 mo postpartum they still possessed an aver-
Energy andprotein intakes age of 3.2 kg (range, -5.2-10 kg) more body fat than
These values were calculated from a record kept by each before pregnancy. This last figure should be compared
woman during four consecutive days, including a Saturday or with an average remaining weight gain of < 1 kg 6 mo
Sunday, that were typical of her life. The subject was supplied postpartum.
with a balance that weighed to the nearest gram and given writ- Table 2 also shows that RMR increased throughout
ten and oral instructions on how to weigh all foodsand measure pregnancy; the average figure in week 36 was 130% of the
all drinks consumed. The representativeness ofthe food habits prepregnancy value. RMR expressed as percent of the
during these days for each woman was checked by a 1-2 h per- corresponding prepregnant value showed a high stan-
sonal interview about her food habits. The energy and protein
dard deviation at all measurement times, which indicates
contents of the foods recorded were calculated using Swedish
food tables(13). that the response of an individual woman could be very
different from the average presented in the table.
Statistical analyses The mean values (±SD) ofthe cumulative increase in
Statistical treatment of the data (regression analysis and t RMR for the mothers for the first (before week 21) and
test of paired observations) was performed with conventional second (after week 21) halves of pregnancy and for the
methods ( 14). complete pregnancy were 1 1 600 ± 21 700, 40 100
± 24 000, and 50 300 ± 37 400 kcal, respectively. Re-
suits ofregression analyses between cumulative increase
Results
in RMR and birth weight ofthe baby (BBW) are shown
MBW, TBW, TBK, TBF, and RMR measurements in Table 3. (For regression analysis RMR [in kilocalo-
from before and during pregnancy as well as postpartum ries] was x and BBW [in grams] was y.) The complete
are shown in Table 2. Although these women increased increase in RMR during pregnancy for this group of
their body weights during early pregnancy their TBK de- women (50 300 kcal) had a very large SD (37 400 kcal).
creased significantly. Also, 6 mo postpartum TBK was The increased metabolic rate during the second half of
significantly lower although MBW was slightly higher pregnancy was responsible for most (80%) of the total
compared with the prepregnant value. Similarly, TBW increase. Regression analysis showed that the cumulative
 ENERGY REQUIREMENTS DURING PREGNANCY 945

TABLE 3
Correlation between cumulative increase in RMR and birth weight of baby’

Slope of Level of
regression line Intercept r sigificance of r

1st halfofpregnancy 5.2 X l0 3505 0.22 NS

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

2nd halfofpregnancy 9.52 x l0 3176 0.51 0.05 >p> 0.01
Complete pregnancy 5.29 x i0 3292 0.44 0.05 > p > 0.01

#{149}n22

increase in RMR during the second half of pregnancy any ofthe measurements. However, at the first measure-
and during the complete pregnancy correlated signifi- ment during pregnancy the correlation between the
cantly with BBW whereas no such significant correlation change in TPK and BBW in relation to MH (BBW:MH)
was found in the first halfofpregnancy. The total cumu- was significant. Changes in MBW were significantly cor-
lative increase in RMR during pregnancy was not corre- related with changes in TBW 5-10 d postpartum and
lated with nutritional status, expressed as body mass in- with changes in TBF 6 mo postpartum. A significant cor-
dcx or percent body fat, or with energy intake of the relation, not shown in the table, was found between
mother before pregnancy. changes in MBW and RMR 6 mo postpartum. Except
To investigate possible correlations between changes for the relationship between MBW and TBF shown in
in RMR, TBK, MBW, TBF, and TBW, regression analy- the table, changes in TBF were not significantly corre-
ses were performed at each measurement between all lated with changes in MBW, TBP, RMR, or BBW for
possible combinations of the above variables. In these any ofthe measurements. Correlations between changes
calculations the figures used were always expressed as in TBW and TBF at all measurements were significant
percent of the value obtained at the prepregnancy mea- but the results are not shown in Table 4.
surement. Some of the correlations considered to be of
interest are given in Table 4. In addition, regression anal-
ysis between changes in the variables mentioned above Discussion
and BBW and, for TPK, BBW in relation to the maternal
height (BBW:MH) were performed; significant corre- The kind ofstudy reported in this paper cannot be car-
lations are shown in Table 4. ried out on a truly random sample of the population.
Confirming findings presented in Table 3, the relation- Thus, it becomes important to consider how representa-
ship between the increased RMR and BBW appeared to tive our women are for Swedish women in general. The
be related to events occurring during the second half of average increase in weight during pregnancy and the av-
pregnancy because significant correlations between these erage birth weight of the babies of our subjects are very
variables were obtained at gestational weeks 30 and 36 similar to corresponding figures compiled from a popu-
but not at weeks 16-18. Note that the changes in TBK lation of 444 women at Huddinge University Hospital
and RMR were correlated at the second measurement during recent years, suggesting that it is reasonable to re-
during pregnancy but not at the other measurements. gard our group of women as fairly representative of the
The changes in TBK were not significantly correlated population.
with changes in MBW, TBW, or TBF or with BBW at The results ofthis study differ in several respects from

TABLE 4
Results ofregression analyses between RMR, MBW, TBK, TBW, TBF, BBW, and BBW in relation to maternal height (BBW:MH) during and
after pregnancy

Slope of Level of
x y Time regression line Intercept r significance of r

RMR TBK Gestational weeks 16-1 8 0.277 63.7 0.4 1 p > 0.05
RMR TBK Gestational week 30 0.239 66.2 0.48 0.05 > p > 0.01
RMR BBWt Gestational week 30 14.3 1793 0.58 0.01 > p > 0.001
RMR BBWt Gestational week 36 10.5 2 167 0.45 0.05 > p > 0.01
TBK BBW:MHt Gestational weeks 16-1 8 0. 1 13 1 1.0 0.47 0.05 > p > 0.01
MBW TBW 5-10 d postpartum 1. 10 - 18.5 0.49 0.05 > p > 0.01
MBW TBF 6 mo postpartum 3.00 - I 82 0.60 0.01 > p > 0.001

a Values used in the analysis were percents ofcorresponding prepregnancy values except where indicated otherwise.
t g.
g/cm.
946 FORSUM ET AL

results obtained in similar studies ofwell-nourished preg- decrease due to pregnancy could not be demonstrated.
nant women. For example, in studies on Scottish (15) The unexpected results with respect to TBK obtained in
and Dutch (16) women the amount of fat gained during this study raise some important questions about whole-
pregnancy was found to be considerably smaller than the body counting. Our calibration procedure took into ac-
figure obtained in our study. This could of course mdi- count that subjects have different heights and weights but
cate a true difference but at least part of the difference apart from that no special attempts were made to correct

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

could be due to many ofthe Dutch and Scottish women for the changes in body shape occurring during preg-
being examined for the first time around week 10 of preg- nancy. It is possible that a changing geometry during
nancy. According to our study a considerable amount of pregnancy influenced our results although it seems un-
fat is retained early in pregnancy and it is thus possible likely that such effects could have been very important
that a significant part ofthe retained fat is present already in the beginning of pregnancy or after delivery. Another
at week 10 ofpregnancy. Another possible reason for the factor that could be important was the increased amount
different results is that different techniques for the esti- ofsubcutaneous fat retained by the women, which might
mation ofbody fat were used. In the Scottish and Dutch have attenuated the radiation from K in the cells.
studies underwater weighing and skinfold measurements Our findings ofa decrease in TBK during the first part
were used. ofpregnancy and 6 mo postpartum need to be confirmed
We used a model developed by Pipe et al ( 10) for calcu- in other studies because of the uncertainties mentioned
lating TBF in the maternal body. We chose this model above. In nonpregnant adults a decrease in lean body
because it is based on measurements ofTBK, TBW, and mass would be the obvious interpretation of such find-
MBW. The results of our K measurements were some- ings and it is possible that this is at least part ofthe expla-
what unexpected and suggest that some of the funda- nation because TBW showed a tendency to decrease in
mental assumptions made in the model may not be the beginning ofpregnancy and 6 mo postpartum. How-
valid. However, estimates ofbody fat based on measure- ever, changes in TBK and TBW were not correlated, in-
ments of body water and body weight only (17) corre- dicating that some other factor (or factors) is responsible
lated well with estimates based on the model given by for the decrease in TBK. In a recent paper Cheek et al
Pipe et al (10) (r varied between 0.98 and 0.99 at the (18) reported that muscle samples from pregnant women
different measurements). The construction ofthe model contained less intracellular K and water than those from
is such that when the body weight ofthe subject increases nonpregnant women, a finding that supports our results.
and the body water decreases, which was the situation Our finding that the change in TBK was related to the
for several subjects at the first pregnancy measurement, change in energy metabolism during the first part of
the increase in body fat will be larger than the increase in pregnancy supports the idea that the change in TBK is
body weight. According to current concepts such the result of some physiological event. A possible hy-
changes occur when the composition of the body pothesis to explain our findings is that the activity of the
changes, ie, when the concentration of fat in the body sodium-potassium pump, which is often assumed to
increases. However, these concepts are based on an as- make up a considerable part of the resting metabolism,
sumption about the water content of the fat-free body may vary in response to the demands of pregnancy on
that is not necessarily valid during pregnancy and post- the energy metabolism of the woman. Data in support
partum. Our results ofbody fat estimations should thus of this hypothesis were presented recently (19) although
be considered somewhat tentative and they need to be different opinions appear to exist regarding the potential
compared with other independent methods for body-fat of the Na-K pump for saving energy (20, 21). Another
estimations. We find this especially important because in possibility is that the decreased K content is related to
our study increases in body weight correlated poorly with the well-known changes in water and electrolyte balances
increases in body fat. during pregnancy. Hear et al (22) studied the variability
The values for TBW reported in this paper are based in water and electrolyte contents in human skeletal mus-
on the analysis of 180 in saliva samples obtained before cle and their results indicate that the contents ofNa and
and 4 h after dosing. However, for 14 measurements K in this organ are variable and strongly negatively corre-
(three prepregnant, five during pregnancy, and six post- lated. Finally, however, the K content ofthe different or-
partum) the urine samples collected were also analyzed gans ofrat dams was not affected by pregnancy except in
and the 180 enrichment of urine at time zero was calcu- the case ofthe skeleton, where a significant decrease was
lated with single-pool kinetics (7). The difference be- found (23).
tween the 180 enrichment in saliva and urine varied from An interesting result of this study is that the cumula-
0 to 3% of the enrichment of saliva. Based on these re- tive increase in RMR correlated with BBW and that this
sults we considered analyses ofsaliva samples before and relationship appears to be a feature of the second half
4 h after dosing appropriate for estimating TBW during of pregnancy. This fits well with results from Gambian
pregnancy and postpartum. women (3) who increased their RMR and gave birth to
The finding that TBK decreased significantly during heavier babies after energy supplementation during preg-
the first part ofpregnancy was unexpected. Pipe et al (10) nancy. Our study, in agreement with earlier findings
also examined TBK in pregnant women serially but, be- (24), does not substantiate any relationship between ma-
cause prepregnancy measurements were not included, a ternal energy stores and BBW, although a relationship
 ENERGY REQUIREMENTS DURING PREGNANCY 947

was found between BBW:MH and the change of mater- 8. Reizenstein P. Clinical whole body counting. Bristol: John Wright
nal body K in the beginning of pregnancy. This is in and Sons, Ltd. 1973.
agreement with a suggestion by Briend (25) that a factor 9. Ericsson F, Carlmark B, Eliasson K. Erythrocyte and total body
potassium in untreated primary hypertension. Acts Med Scand
associated with maternal lean body mass rather than
198 1;209:439-44.
with maternal energy stores is involved in the regulation
10. Pipe NGJ, Smith T, Halliday D, Edmonds Ci, Williams C, Coltart
of BBW. This raises the question of the significance of TM. Changes in fat, fat-free mass and body water in human normal

Downloaded from https://academic.oup.com/ajcn/article-abstract/47/6/942/4694880 by University of New England user on 11 January 2019

maternal energy stores and their regulation. pregnancy. Br J Obstet Gynaecol 1979; 86:929-40.
The women in this study gained more weight and body 1 I . Scholander PF. Analysis of accurate estimations of respiratory
fat, delivered heavier babies, and increased their RMR gases in one-halfcubic centimeter samples. J Biol Chem 1947; 167:
more during pregnancy than the figures by Hytten and 235-50.
Leitch (1) suggest as the average for well-nourished popu- 12. Brouwer E. Report of sub-committee on constants and factors.
lations ofwomen. However, it is not possible to conclude London: Academic Press, 1965. (European Association for Animal
that either ofthese two sets offigures is more representa- Production publication #1 1.)
13. The National Food Administration. Livsmedelstabeller. Stock.
tive of an optimal nutritional situation than the other.
holm: Liber Tryck, 1978.
Also, all estimates related to energy requirements show
14. Armitage P. Statistical methods in medical research. New York:
large SDs, which demonstrates the difficulties in defining Halsted Press, John Wiley & Sons Inc. 1971.
a reference woman on whom the estimates ofenergy re- 15. Durnin JVGA, McKillop FM, Grant S. Fitzgerald G. Energy re-
quirements during pregnancy could be based. We are ap- quirements ofpregnancy in Scotland. Lancet 1987;2:897-900.
parently far from being able to estimate energy require- 16. van Raaij JMA, Vermaat-Miedema SH, Schonk CM, Peek MEM,
ments during human reproduction. 13 Hautvast JGAJ. Energy requirements of pregnancy in the Nether-
lands. Lancet l987;2:953-5.
We thank all the women who participated in this study for their
1 7. Moore FD, Olesen KH, McMurrey JD, Parker HV, Ball MR, Boy-
devoted cooperation. We also wish to express our gratitude to Profes-
den CM. The body cell mass and its supporting environment. Phil.
sors B Pernow and L Kaiser at the Department ofClinical Physiology,
adelphia: WB Saunders Co. 1963.
Karolinska Hospital, for help with measurements of RMR and to Dr
18. Cheek DB, Petrucco OM, Gillespie A, Ness D, Green RC. Muscle
P Reizenstein and Ms M Ekman for help with measurements of TBK. cellgrowth and the distribution ofwater and electrolytes in human
MsI Nilsson, Ms M Thil#{233}n,and Ms B H#{228}ggkvistare gratefully acknowl-
pregnancy. Early Hum Dev 1985; 1 1:293-305.
edged for skillful technical assistance. We thank Dr F Westman for his 19. Milligan LP, McBride BW. Shifts in animal energy requirements
help with isotope analyses. across physiological and alimentational states. Energy costs of ion
pumpingbyanimaltissues. J Nutr l985;1l5:l374-82.

References 20. Biron R, Burger A, Chinet A, Claussen T, Dubois-Ferri#{233}re R. Thy-

roid hormones and the energetics of active sodium-potassium
1. Hytten FE, Leitch I. The physiology ofhuman pregnancy. 2nd ed. transport in mammalian skeletal muscles. J Physiol l979;297:47-
Oxford: Blackwell Scientific Publications, 1971. 60.
2. Thomson AM. Diet in pregnancy. 3 Diet in relation to the course 21. Chinet A, Clausen T. Energetics ofactive sodium-potassium trans-
and outcome ofpregnancy. BrJ Nutr 1959; 13:509-25. port following stimulation with insulin, adrenaline or salbutamol
3. Lawrence M, Lamb WH, Lawrence F, Whitehead RG. Mainte- in rat soleus muscle. Pflugers Arch l984;40l: 160-6.
nance energy cost ofpregnancy in rural Gambian women and in- 22. Hear CTG, Carpenter RG, florence I. Variability in the water,
fluence ofdietary status. Lancet l984;2:363-5. sodium, potassium and chloride content ofhuman skeletal muscle.
4. Forsum E, Sadurskis A, Wager J. Energy maintenance cost during J Gin Pathol 1965; 18:74-81.
pregnancy in healthy Swedish women. Lancet 1985; 1:107-8. 23. Romeu A, Acola L, Alemany M. Sodium, potassium, magnesium
5. Sadurskis A, Kabir N, Wager J, Forsum E. Energy metabolism, and calcium tissue levels in the rat during pregnancy. Biol Res
body composition and milk production duringlactation in healthy Pregnancy l986;7:52-5.
Swedish women. Am J Clin Nutr(in press). 24. Briend A. Normal fetal growth regulation: nutritional aspects. In:
6. Schoeller DA, van Santen E, Peterson DW, Dicta W, Jaspan J, Grasey M, Falkner F, eds. Nutritional needs and assessment of nor-
Klein PD. Total body water measurements in humans with 80 mal growth. New York: Nestl#{233} Nutrition Vevey Raven Press, 1985:
and 2H labeled water. Am J Clin Nutr 1980; 33:2686-93. 1-21.
7. Shipley RA, Gark RE. Tracer methods for in vivo kinetics. Theory 25. Briend A. Do maternal energy reserves limit fetal growth? Lancet
and applications. New York: Academic Press, 1972. 1985; 1:38-40.