Previous - Next
J.V.G.A. DURNIN*
* Institute of Physiology, University of Glasgow, Glasgow G12 8QQ, Scotland, U.K.
1. Introduction
2. Background
3. Methodology
4. Results and discussion
5. Conclusion on pregnancy data
6. Lactation
References
This paper is a summary of
the study funded by the Nestlé Foundation on the energy
requirements of pregnancy (and lactation) which has been in
progress in five countries during a period of about five years.
The countries are Scotland, Holland, The Gambia, Thailand and the
Philippines.
Clearly, neither Scotland nor Holland would come into the category of countries where energy deficiency was likely to occur in any prevalence. They seem therefore outside the remit of this workshop. Moreover, even in the three developing countries, the populations were chosen to be representative of groups where obvious malnutrition was improbable.
Thus, this whole study, by and large, is being undertaken in environments that would not be expected, on superficial grounds, to produce data of direct relevance to our immediate purposes. However, the findings do indeed have considerable practical importance since they have been obtained on two groups of adults who constitute a major proportion of those likely to benefit nutritionally from food supplementation - i.e., pregnant and lactating mothers. There is obviously a direct usefulness in having much more definitive and wide-ranging longitudinal information on the real-life extra energy needed by pregnancy and lactation which could easily have a bearing on the final conclusions and recommendations of this workshop.
It should be noted that although the Nestlé Foundation project on energy requirements initially included lactating as well as pregnant women, it became clear after some of the women had been studied that the breadth and intensity of the experimental measurements required by pregnancy alone were such that only limited information could be collected strictly on lactation. Thus, although in The Gambia fairly extensive measurements were done on some aspects of lactation, in Scotland, Holland and in Thailand, only tentative conclusions can be drawn from the lactation data, and no real attempt was made to study lactation in the Philippines.
As will
become apparent, much of the information obtained in these five
countries is consistent enough to provide a solid baseline from
which relevant deductions can be made for present purposes.
Pregnancy and lactation are
both conditions where the average requirements for energy of a
woman might be expected to be increased considerably. It is
possible to calculate, with substantial theoretical accuracy, the
extra energy required by the mother to increase her own tissues
and fluids appropriately, and to produce a normal healthy baby,
and, after the baby is born, to breast-feed it for six months or
longer (HYTTEN and LEITCH, 1971; FAO/WHO/UNU, 1985). These
theoretical calculations have been more or less accepted in the
FAO/WHO/UNU Report as representing the practical situation. The
result is that the Report recommends that pregnant women who
remain on "full activity" - which presumably means that
there is no change in the way of life, nor a change of a
biological nature - need an extra 285 kcal/d, and
lactating women (either in the first 6 months, or after 6 months)
need an extra 500 kcal/d.
These represent quite considerable additions to the average diet of a non-pregnant, non-lactating woman. Thus, in areas where chronic energy deficiency exists, and where food availability is therefore restricted, the implication is that either physical activity needs to be quite markedly reduced, or supplementation needs to be provided on an extensive scale. In fact, not only would physical activity need to be diminished proportionately to the energy-saving required, but the diminution would be disproportionate, since the heavier body mass of the pregnant woman means that everything she does - from sleeping to housework, walking, working in the fields - would use up more energy than in the pre-pregnant state when body mass is less. Such a decrease in activity places a burden not only on the woman but on the members of the family who receive less help and attention. If the extra requirements of energy need to be met in full, they become quantitatively probably by far the most important demand on any allocation of food supplements available for a community.
It is therefore a matter of some significance to know whether or not pregnancy and lactation really require these considerable extra quantities of energy, particularly in the light of the comparative dearth of well-controlled experimental studies, together with the conflicting nature of some of the data.
The study
which is described here was planned as an attempt to obtain
wide-ranging information of a longitudinal nature on several
contrasting groups of healthy mothers, living in both developed
and developing countries.
3.1. Selection of subjects
3.2. Body weight and body fat
3.3. Energy intake
3.4. Basal metabolic rate (BMR)
3.5. Standardized exercise test
3.6. Normal daily activity pattern
3.7. Daily energy expenditure
3.8. Frequency of measurements
The studies were carried
out in five different countries: 1) on a socioeconomically poor
and a middle-class urban group in Glasgow, Scotland; 2) a
middle-class urban community in Wageningen, Holland; 3) poor
rural women in The Gambia; 4) rural women in Thailand, and 5)
women living in a rural area in the Philippines. The groups of
women living in the developing countries were chosen on the basis
that they were not malnourished, and this was universally true
with the possible exception of some of the women in The Gambia
where seasonal fluctuations in food availability might
occasionally have had some influence.
In order to try and eliminate as many extraneous confusing influences as possible, the selection of the different populations of women, and the methodology, were standardized as far as practicable. The initial plan was that 40-50 women were to be selected in each centre; in the end, some centres recruited more than this number. In the two developed countries it was thought desirable that the women should be full-time housewives since if they had another occupation they would have to give it up at some stage in pregnancy and this might clearly upset the previous normal routine of everyday life, particularly in relation to levels of physical activity.
The women in the developed countries had to have the firm intention of breast-feeding, since otherwise this might influence the desirable weight and fat gain, and therefore it was preferable that they should be in their second pregnancy so that they should already have had an experience of breast-feeding. The age range was between 20 and 30 years. Obviously it was necessary, as far as possible, that the women should be healthy and normal from their medical and reproductive history, they should not have any potentially complicating attitudes to diet, etc., and they should neither be very fat nor very thin.
On the whole, most of these requirements were met with the exception that, in the developing countries, many of the mothers had already had more than one previous pregnancy and, of course, almost all of them continued to work in the home and in the fields right up until term.
In all the centres, considerable effort was made to recruit the mothers as early as possible in pregnancy and preferably before 810 weeks gestation, although again this was more difficult ;n Thailand and in the Philippines. Because of special circumstances in The Gambia, many of these women were actually studied initially before they became pregnant.
In order to
standardize the measurements in the very different populations
and circumstances, the senior field worker from each centre (with
the exception of the Philippines) spent some weeks or months in
Glasgow either learning or renewing experience of the various
techniques. In the end, a very high degree of uniformity was
achieved. A variety of measurements on all of the pregnant women
in each of the five centres were carried out to try and cover all
of the different variables which might influence energy
requirements. These were 1) body weight; 2) body fat to assess
the amount of extra fat normally deposited during pregnancy; 3)
energy intake in the diet; 4) BMR, since the extent of the
increase in BMR can have a marked effect on the energy needs; 5)
possible changes in the efficiency of muscular movement by a
standardized exercise test (walking on a treadmill at fixed
speed); 6) the pattern of normal daily activity was assessed,
using the activity-diary technique, to see whether alterations in
the frequency and/or duration of various activities had taken
place; and 7) an estimate of total daily energy expenditure was
attempted using the activity-diary combined with indirect
calorimetric measurements.
The women in all centres
were weighed on calibrated beam balances on each visit to the
laboratory (intervals between visits in most centres were
6-weekly). In Scotland and in Holland each woman was also given a
high-quality bathroom scale which she kept at home and which was
frequently calibrated; all of these mothers weighed themselves
weekly, and about half the total number recorded their weight
daily.
Body fat was assessed in all centres using 4 skinfold thicknesses as described by DURNIN and WOMERSLEY (1974), although in the later stages of pregnancy it became increasingly difficult to pick up the supra-iliac skinfold, and in these cases the equation using 3 skinfolds to predict fatness was employed. In Scotland and in Holland the underwater densitometric method was also used on some women. In The Gambia total body water, using deuterium, was also used as an estimate of fatness.
Two slight variants of these techniques were employed in Scotland and in Holland. In the Glasgow women, skinfold thicknesses were used as an estimate of total body-fat equivalent to that which would be present at the end of pregnancy by measuring the skinfolds two weeks after the delivery of the baby. At this stage, it appears that the body composition of the women has more or less returned to the prepregnant state (with the exception of extra breast tissue and the added maternal fat) and therefore the equations used to predict body fat from skinfold thicknesses should be reasonably valid.
In Holland,
adjustments were made to the density of the fat-free mass at
varying stages of pregnancy by calculating the change in body
composition of the fat-free mass which results from the altered
body fluids and tissues of the pregnant women.
Because this was probably
the most important single measurement, the method chosen was the
one usually accepted as being the most accurate in the
circumstances. In Scotland and in Holland this was the individual
weighed inventory method: each item of food or drink was weighed
and recorded immediately before consumption. Leftovers were also
weighed. The duration of each period of measurement was five
consecutive days. In Thailand and the Philippines, the
measurement was done by the precise weighing method: raw weights
of food were obtained prior to cooking, as well as the cooked
weights. In The Gambia, because of special local problems, the
data on energy intake were thought to be subject to some doubt
and not included in the results. In the other four centres, the
energy value of the edible portion of the foods actually consumed
was calculated using tables of food composition appropriate to
the particular region.
Alterations in BMR could
account for a considerable proportion of the total theoretical
cost of pregnancy. It has also been suggested that adaptations in
BMR might result in energy savings. In each centre, BMR was
measured using the Douglas Bag technique of indirect calorimetry.
The women were studied in the resting, fasted state first thing
in the morning and, after half to three-quarters of an hour of
lying in bed in a quiet environment, two measurements were done,
the first lasting 15 minutes and the second, 10.
The volume
of expired air was measured using a calibrated air meter, and the
O2 and CO2 contents of the expired air were
obtained by the Servomex 570A paramagnetic oxygen analyser and
the PK Morgan infra-red CO2 analyser, both of these
instruments also being frequently calibrated against known gas
mixtures. Agreement for the duplicate measures of BMR was
expected to be within 3%; if this was not the case, a third
measurement was done and the mean value of all three was taken as
the BMR. BMR was measured in the laboratory, to obtain quietness
with a minimum of extraneous distractions.
This was carried out by
having each woman, except in the Philippines, walk at a fixed
pace on a treadmill. The rate was 3.8 km/h on a level gradient
for the women in Scotland, in Holland and in The Gambia, and 3.0
km/h for the women in Thailand and in the Philippines. After one
or more initial periods to allow the woman to become accustomed
to the exercise, there was a 3-minute "run-in" period
when she breathed through the mouthpiece valve before collection
of expired air was begun. This was done into a Douglas Bag for
two separate 10-minute periods. Analyses were done as for the
BMR. A variation of more than 5% between samples meant that a
third measurement was done. The mean of all the readings was
taken as the exercise metabolic rate.
A record of the activity
pattern of each mother was obtained for the whole 24 hours on
each of five consecutive days. This was done using the
activity-diary technique first described by DURNIN (1952),
whereby a detailed minute-by-minute record was procured of all
the separate activities of the day. These were then subdivided in
large categories - bed, sitting, standing, housework, walking -
so that comparisons could be made of any changes in the total
time spent in these activities as pregnancy progressed.
The
activity-diary was filled in by the mothers themselves in
Scotland and in Holland; in the three developing countries this
was done by a trained (local) observer.
This was calculated from
the data on the duration of the single activities, combined with
measurements by indirect calorimetry (using the Douglas Bag) of
the energy cost (in kcal/min) of each of the important activities
(c/f DURNIN and BROCKWAY, 1959). Great care was taken to try and
obtain records of activity and of energy cost which appeared to
be as near the "normal" as possible for each woman.
Energy expenditure, as a total for the whole day, was measured on
the five consecutive days of the activity-diary record.
There were some differences
in the frequency with which the various measurements were made in
the five countries. The relevant information for body weight and
body fat has already been given. For energy intake, BMR, activity
pattern, and daily energy expenditure, the norm in all centres
was that all these measurements were carried out at 6-weekly
intervals. In Glasgow, after the first 21 women, these
measurements were made either at 4-weekly intervals, or at
2-weekly intervals. This was done because of the large
variability, both within as well as between individuals,
particularly for energy intake and BMR.