Free-living techniques
energy expenditure measured by two independent in pregnant and nonpregnant Gambian women
A. HEINI, Y. SCHUTZ, E. DIAZ, A. M. PRENTICE, R. G. WHITEHEAD, AND E. JRQUIER Institute of Physiology, Faculty of Medicine, University of Lausanne, CH-1005 Lausanne, Switzerland; and Medical Research Council, Dunn Nutrition Unit, Keneba, The Gambia; and Cambridge CBA IXJ, United Kingdom A., Y. SCHUTZ, E. DIAZ, A. M. PRENTICE,R. G. AND E. J~QUIER. Free-Ziuing energy expenditure measured by two independent techniques in pregnant and nonpregnant Gambian women. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E9-E17, 1991.-Free-living energy expenditure (EE) was assessed in 37 young pregnant Gambian women at the 12th (n = 11, 53.5 t 1.7 kg), 24th (n = 14, 54.7 t 2.1 kg), and 36th (n = 12, 65.0 t 2.6 kg) wk of pregnancy and was compared with nonpregnant nonlactating (NPNL) control women (n = 12, 50.3 t 1.6 kg). The following two methods were used to assess EE: 1) the heart rate (HR) method using individual regression lines (HR vs. EE) established at different activity levels in a respiration chamber and 2) the doubly labeled water (2H2180) method in a subgroup of 25 pregnant and 7 control women. With the HR method the EE during the agricultural rainy season was found to be 2,408 t 87, 2,293 t 122, and 2,782 t 130 kcal/day at 12,24, and 36 wk of gestation and were not significantly different from the control group (2,502 t 133 kcal/day). These findings were confirmed by the ‘H2180 measurements, which failed to show any effect of pregnancy on EE. Expressed per unit body weight, the free-living EE was found to be lower (P < 0.01 with ‘Hz”0 method) at 36 wk of gestation than in the NPNL group. It is concluded that, in these Gambian women, energy-sparing mechanisms that contribute to meet the additional energy stress of gestation are operating during pregnancy (e.g., diminished spontaneous physical activity). HEINI, WHITEHEAD,
energy metabolism; physical activity
heart rate; doubly labeled water; pregnancy;
PREGNANCYAND LACTATION are two physiologicalproc-
esses of anabolic nature that increase the energy requirement of the mother (13, 14). In pregnancy the energy requirements are increased for two reasons. First, there is a rise in energy expenditure (EE) due to an increase in basal metabolic rate and a higher cost of movernent by the pregnant woman who gains weight. Second, there is an enhanced energy storage both of fetal and maternal origins (14). It is, however, not known whether these energy requirements must be wholly met by an increased food intake or whether adaptive mechanisms can spare energy during pregnancy. The assessment of 24-h EE in free-living populations of Third World countries is of great importance for two main reasons. First, in its latest report, the FAO/WHO/ UNU (13) expert committee recommended that energy requirements of adults should be assessed on the basis 0193-1849/91
$1.50 Copyright
of EE rather than energy intake data. Second, women living in Third World countries not only appear to be more physically active than those in industralized countries, but they are also pregnant more frequently and lactate longer, despite a level of food intake that has only been shown to be chronically marginal (23) by international standards (12, 13). The purpose of the present study was to assess EE in free-living conditions in pregnant and nonpregnant women living in a rural village of The Gambia. Two independent methods were used to assess EE [heart rate (HR) and doubly labeled water (2H2’“o)] to find out whether adaptive mechanisms to the marginal energy intake (23) occur during pregnancy. MATERIALS AND METHODS
The study was carried out in women from three rural Gambian villages (Keneba, Manduar, and Kantonkunda) of the subsahelian region. A detailed description of the environmental conditions prevailing in these villages has been previously published (16). The study period covered the beginning of the rainy agricultural season (July, 1987) to the dry season (April, 1988). The rainy season is characterized by heavy farming work and food shortage, whereas, during the more favorable dry season, the food availability is increased. Thirty-seven young pregnant women and twelve matched nonpregnant nonlactating (NPNL) women, who served as controls, took part in this crosssectional study. The pregnant women were studied at 12 (n = ll), 24 (n = 14), and 36 (n = 12) wk of pregnancy. Pregnancy was excluded or confirmed by a pregnancy test. Assessments of gestational age at birth were based on Dubowitz scores (8) by retrospective calculations. The women were apparently healthy, and none had complications or presented any clinical signs of malnutrition. The physical characteristics of women are shown in Table 1. The women gave birth to babies whose weight averaged 3.05 t 0.45, 2.75 t 0.29, and 2.96 t 0.26 kg for the women studied in the first, second, and third trimesters of pregnancy, respectively. Acceptance to participate in the study was obtained after giving a detailed explanation of the procedure to be used. The study was approved by the Medical Research Council/Gambian Governmental Ethical Committee. Anthropometry and body composition. Body weights were measured with a platform digital scale that was
0 1991 the American
Physiological
Society
E9
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
El0
ENERGY
TABLE
1. Physical characteristics Subject
Age,
1 2 3 4 5 6 7 8 9 10 11 12 Means&SE
22 37 24 25 26 32 24 32 26 24 30 29
Yr
EXPENDITURE
IN
PREGNANT
GAMBIANS
of pregnant and nonpregnant no&acting women
Weight, kg
Height, cm
BMI, kg/m2
Height2
Body Fat, % wt
FFM, kg
29.9
20.0 28.7 18.7 16.5 22.6 17.2 21.5 29.5 19.7 18.9 26.3 22.5t1.4
39.1 24.1 37.6 34.6 36.4 40.9 40.1 43.8 35.9 36.6 47.0 40.6 38.9tl.l
45.5 36.4 38.4 40.5 38.3 41.1 43.1 57.3 36.1 35.7 49.6 41.0 41.021.8
21.9 24.8 19.5 25.4 26.5 24.4 20.2 20.3 28.7 21.2 19.4 22.9t1.0
38.8 39.9 39.5 39.7 45.1 42.8 39.3 37.1 46.1 43.4 39.9 41.1t0.9
43.1 46.6 41.0 48.0 43.1 40.3 35.7 49.4 49.3 40.0 43.7t1.5
18.8 23.5 23.0 24.5 32.3 26.5 19.5 20.9 22.1 23.1 20.4 30.3 22.8
39.0 44.4 50.4 37.6 39.5 42.4 42.3 38.2 42.6 41.5 41.1 35.7 49.8 38.3 41.6t1.5
39.3 46.0 52.0 35.2 44.8 43.7 51.1 43.6 49.3 45.0 46.4 37.7 44.7 43.9 44.5k1.3
Body
Resistance,
cm2/C!
NPNL
27.6t1.3
55.8 42.6 52.8 42.6 43.6 52.8 48.4 55.8 51.0 45.6 58.0 55.1 50.3t1.6
153.0 156.0 162.5 150.0 163.0 155.0 161.0 167.0 153.5 154.0 168.0 160.0 158.5t1.5
49.7 53.0 49.1 53.2 61.3 56.6 49.3 46.6 64.7 55.0 49.5 53.5t1.7
155.0 155.5 158.0 159.0 166.0 167.5 165.5 152.5 165.0 155.5 161.0 160.1t1.6 156.0 160.0 172.0 168.0 162.5 159.5 160.5 154.0 163.0 159.0 157.5 151.5 157.0 156.0 159.8t1.4 155.0 164.0 163.0 179.0 164.0 155.5 160.5 163.0 169.0 161.5 167.0 156.5 163.1t1.9
23.8 17.5 20.0 18.9 16.4 21.9 18.7 20.0 21.5 19.2 20.5 21.5 20.0t0.6 1 st Trimester
13 14 15 16 17 18 19 20 21 22 23 Means&SE
20 28 22 23
18 25 25 30 32
21 26
24.6t1.3
20.7 21.9 19.7 21.0 22.2 20.2 18.0 20.0 23.8 22.7 19.1 20.9t0.5 2nd Trimester
24 25 26 27 28 29 30 31 32 33 34 35 36 37 MeanstSE
28 25 32 17 35 26 22 35 22 21 28 22 26.1t1.4
48.0 58.0 65.5 47.2 52.3 62.6 57.6 47.5 53.9 53.3 53.4 44.9 71.4 49.6 54.7t2.1
38 39 40 41 42 43 44 45 46 47 48 49 MeansHE
20 19 31 23 21 20 28 34 22 23 29 33 25.3t1.6
58.0 66.6 64.3 81.1 62.3 57.1 60.7 60.1 72.1 80.0 66.0 51.9 65.0&2.6*
25 27
19.7 22.7 22.1 16.7 19.8 24.6 22.4 20.0 20.3 21.1 21.5 19.6 29.0 20.4
21.4t0.8
23.4tl.l
3rd Trimester
NPNL,
nonpregnant
nonlactating;
BMI,
body
mass index;
FFM,
24.1 24.8 24.2 25.3 23.2 23.6 23.6 22.6 25.2 30.7 23.7 21.1 24.420.7t fat-free
accurate within 50 g. The subjects were under fasting conditions and were wearing a typical African cloth, the weight of which (-0.3 kg) was subtracted from the total body weight. Height was measured with a stadiometer, the accuracy of which was within 0.5 cm. Body composition was assessed using the following
22.3 45.1 46.0 31.7 45.5 50.6 25.5 47.9 46.9 25.8 60.2 57.4 23.6 47.6 52.3 20.6 45.3 40.8 19.8 48.7 45.8 21.0 47.5 54.3 26.7 52.9 58.3 32.2 54.2 64.7 21.8 51.6 51.4 25.0 38.9 46.8 24.7tl.l 48.8k1.61 51.3+1.9? mass. * P < 0.01, t P < 0.05, and $ P < 0.001 vs. other groups.
two independent methods: 1) skinfold thicknesses measured at four sites (biceps, triceps, subscapular, and suprailiac) (10) and 2) whole body resistance (20) measured at frequencies of 0.5, 50, and 100 kHz with a portable device developed at the Institute of Physiology (unpublished observations). For each subject, height squared
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
ENERGY
EXPENDITURE
IN
GAMBIANS
El1
all periods in the calorimeter during which the daytime HR was below this critical HR rate, an average value of sedentary EE was calculated for each subject (Fig. 1). This average value was used in everyday life situations to compute EE when HR was below the critical value. Leg work predominated in the treadmill exercises that were used to establish the relationship between EE and HR in the respiration chamber. However, visual assessment of the work done by the women in the fields shows that physical activity involved both arm and leg work. To ascertain whether the regression line was influenced by the nature of the work performed, a different group of subjects (9 NPNL women aged between 19 and 39 yr) was studied; they mimicked the type of activities performed in the field while in the respiration chamber. The exercise consisted in moving a l-kg block of wood on a horizontal surface and lifting it to two different heights (15 and 30 cm) each, over a lo-min period. In addition to the exercise period, two sitting values were measured over 10 min before exercise. For each subject the linear regression lines obtained with this combination of arm plus leg work was compared with that previously obtained with the treadmill exercise. In free-living conditions, HR was recorded during normal daily activities. The investigator went to the subject’s home at 7:30 A.M., placed the electrode belt on the chest, checked HR, and initiated the recording. Similarly, in the evening, the investigator went to stop the recording and to collect the HR monitor at -7:30 P.M. The data were then immediately printed in the laboratory. In each subject, the duration of measurements ranged from 1 to 4 days. The sedentary rate of EE previously described was used to cover the period between 7:30 P.M. and the time when the subject went to bed, typically 2 h later. During the night, HR was not monitored; the nighttime sleeping EE measured in the chamber was used to predict that at home. Calculation of EE. The daytime free-living EE was assessed as follows: after monitoring the minute-by-minute HR during daytime a histogram of HR was generated
divided by whole body resistance at 50 kHz was calculated. (Table 1). Assessment of total EE in free-living conditions. Total daily EE was assessed by two independent methods, a modification of the HR method originally described by Bradfield (4) and the ‘H2180 method in a subgroup of pregnant (n = 25) and NPNL women (n = 7). HR method. This method relies on the relationship between HR and EE and the subsequent measurement of HR in free-living conditions. A respiration chamber was used to assess EE at different activity levels (5, 21). The respiration chamber measures 5.1 x 2.4 x 2.4 m and has a net volume of 27 m3. It was built according to the model previously designed in Lausanne, Switzerland (15). The chamber is maintained at a temperature of 26°C and mixing of the air inside is ensured by an air conditioning system recirculating 1,000 m3 air/h. The performance of the calorimeter was tested using infusions of mixtures of CO2 and nitrogen. The accuracy averaged 99% for O2 and 100% of the “true” value for CO2 with a precision of 4-10 ml/min over a l-h period. HR were recorded by means of one of the following two portable instruments: 1) the “Sport Tester” (Polar electro, Kempele, Finland) and 2) the “Baumann” HR counter (Fleurier, Neuchatel, Switzerland). Both devices are small and light. The HR data were averaged over 1-min intervals and were stored in a memory. To study the relationship between HR and EE, five treadmill exercises of various intensity [speeds 1.4, 1.6, and 2.0 miles/h (mph) on the flat or on slopes of 5 and lo%] were carried out. Each exercise lasted for 15 min. In addition, two sitting values were obtained and a linear least-square regression line between HR, and EE was calculated through the experimental points (Fig. 1). A critical HR was determined below which the relationship between EE and HR was not linear (28). For each individual this value was defined as the midpoint between the HR of the treadmill exercise at the lowest intensity (1.4 mph, flat) and the mean HR obtained during the two sitting periods. A HR below this critical value was considered a “sedentary” level of activity; for 4o .
PREGNANT
[kcal/min]
1 I
3.5 . 3.0 I 2.5 . 2.0 .
1.5 -
EEsed
. ~1,.,111111111111111~~..~~~...~.~ ,,,,,,,,,,,,11,*,,*
1.0 II 0.5 I 0.0 -
I
70
I
Sedentary activity
D
FIG. 1. Individual heart rate (HR) calibration in respiration chamber in 1 woman. HR,,it, critical HR; EE,,d, sedentary energy expenditure; Exe, sitting; Exl-Ex5, treadmill exercise.
: f I
80
3
I
i 90 HR crit
I
I
100
I
I
110
I
I
120 Heart rate [beatslmin]
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
El2
ENERGY
EXPENDITURE
I 54
62
70
78
86
94
102
110
118
126
Heart
5
0 LJ
54
62
70
86
94
102
rate
IJW
110
118
rate
[beets/mlnI
20
0 54
62
70
78
86
94
102
110
118
!26
Heart
2. Typical HR distribution and in free-living conditions
FIG.
(A)
in 1 subject over 2 different
rate
PREGNANT
GAMBIANS
fasting. Serial mid-morning urine samples were collected for the next E-14 days. Isotope enrichment levels were measured using a dual isotope-ratio mass spectrometer (VG Isogas Aquasira, Middlewich, Cheshire, UK). CO2 production was calculated using Coward’s modification (6) of the original Lifson formula (18). The fractionation correction factors used here were 0.99 for 2H20, 1.04 for C802, and 0.93 for H2180. The proportion of water fractionated was calculated for each individual as described elsewhere (24). CO2 production rates were converted to EE using classical indirect calorimetric equations (7a), which require a value for the mean respiratory quotient (RQ) over the measurement period. An assumed RQ of 0.90 derived on the basis of the high-carbohydrate diet consumed in this community was used (2). Propagation of error analysis, as described by Coward et al. (7), was performed on each measurement to obtain an estimate of individual errors. Data analysis. Statistical analyses were performed by using a one-way analysis of variance followed by the multiple comparison test of Tukey. All results presented are expressed as means -+ SE. RESULTS
126
Heart
C
[beats/mln]
IN
[beats/mln]
in respira tion chamber days (B and C).
(Fig. 2). By referring each individual HR to the regression line of HR vs. EE (obtained in the respiration chamber) the daytime free-living EE was calculated. The total EE was obtained by simply summing the three following periods: 1) the daytime EE, 2) sedentary EE before going to bed, and 3) nighttime sleeping EE. ‘&180 method. Total free-living EE was measured using the 2H2’80 method (18). Due to the high cost of isotope, we used this technique only in a subset of women. The principle of the technique together with the full details of experimental method, calculation of results, validation, and limits of error have been described elsewhere (19,24). The subjects received an oral dose of 0.05 g/kg ‘H20 and 0.15 g/kg H2180 after an overnight fast and after the collection of a predose baseline urine sample. Postdose urine samples were collected after a further 4-h eouilibration period during which the subiect was
Antropometry and body composition. The average body weights of the groups of pregnant women were 53.5,54.7, and 65 kg for the first, second, and third trimester groups, respectively, whereas the control group had an average body weight of 50.3 kg (Table 1). The average height of the women in each group was not significantly different from each other. As expected, the average body mass index of the third trimester group was significantly greater than that of the other groups (P < 0.05). In the third trimester the fat-free mass was 25% greater than in the NPNL group when assessed from skin folds (P < 0.001) and when estimated from specific whole body resistance (P < 0.05), respectively (Table 1). HR. A typical HR distribution of one woman obtained in the chamber and in free-living conditions is shown in Fig. 2. Mean daytime HR measured in free-living conditions, resting HR measured in the respiration chamber, and the ratios of these two values are shown in Table 2. Average daytime HR tended to be higher in the pregnant (99-103 beats/min) than in the NPNL women (97 beats/ min), but the difference was statistically significant only in the second trimester (P < 0.05). As expected, resting HR were also higher during pregnancy, but this difference reached statistical significance only between the third trimester (72 t 3 beats/min) and the NPNL groups (61 t 2 beats/min) (P < 0.01). The critical HR was found to be in the order of 84-89 beats/min during pregnancy and 79 beats/min in the NPNL group. The percent of the daytime spent below the critical HR value increased as pregnancy advanced, from 26% in the NPNL group to 33% during the third trimester (Table 2). Regression lines between HR and EE. The average regression characteristics and correlation coefficients of EE vs. HR are shown in Table 3 for each group of women. In the third trimester group. the slope of the regression
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
ENERGY TABLE
2. Heart
IN
PREGNANT
El3
GAMBIANS
rates Mean
Subject
EXPENDITURE
Daily
HR
Critical HR, beats/min
n
Beats/min
1 2 3 4 5 6 7 8 9 10 11 12 Means&SE
2 2 2 2 2 1 2 2 2 2 2 2
97 97 90 94 90 95 110 96 97 99 103 96 97t2
75 86 88 68 68 69 80 88 83 75 84 87 79Ik2
13 14 15 16 17 18 19 20 21 22 23 MeanstSE
2 2 2 2 2 2 2 2 3 2 1
92 94 100 101 98 108 104 105 93 101 95 99t2
78 82 89 79 90 91 73 102 80 77 82 84t3
24 25 26 27 28 29 30 31 32 33 34 35 36 37 MeanskSE
2 3 2 2 2 2 2 2 1 2 1 2 2 2
100 93 94 103 110 105 108 97 109 109 107 103 109 101 103&Z*
96 92 85 94 98 85 81 70 91 78 80 100 93 86 88t2
38 39 40 41 42 43 44 45 46 47 48 49 MeanstSE
2 1 2 2 1 2 3 2 4 2 2 2
95 99 109 97 99 110 103 96 97 101 91 110 101&Z
89 99 86 70 85 99 94 85 87 104 71 94 89t3
%Time Critical
Below HR
Mean
Resting HR, beats/min
Daily
Resting
HR HR
NPNL 12 14 52 25 10 19 22 30 36 24 32 32 261k3
58 71 67 52 54 53 57 73 64 62 50 67 61k2
1.43 1.31 1.27 1.71 1.48 1.64 1.80 1.23 1.52 1.46 1.75 1.50 1.51t0.05
28 26 34 22 32 23 18 30 25 24 35 29t2
55 63 72 62 72 67 63 77 64 62 69 66t2
1.39 1.34 1.22 1.44 1.26 1.42 1.65 1.42 1.37 1.53 1.30 1.40t0.04
47 50 52 31 24 26 18 16 24 18 19 42 32 22 30t3
82 63 68 78 83 62 70 60 64 60 59 76 65 69 69k2
1.19 1.22 1.31 1.12 1.20 1.36 1.46 1.43 1.65 1.47 1.62 1.30 1.70 1.40 1.39t0.05
41 54 22 26 25 22 42 32 29 58 32 15 33t4
68 78 68 62 71 77 79 68 70 89 56 77 72+3t
1.32 1.21 1.42 1.39 1.34 1.38 1.39 1.33 1.35 1.20 1.60 1.38 1.36kO.03
1st Trimester
2nd Trimester
3rd Trimester
HR, heart rate; n, no. of days of measurements. 0.05 and t P < 0.01 vs. NPNL group.
Resting
HR
was the lowest
line was significantly steeper than in the second trimester group (P < 0.05). The slopes of the regression line of EE (kcal/min) on HR (beats/min) were essentially the same when a combination of arm and leg work (0.039 t 0.004) was compared with leg work only (0.039‘t 0.005).
HR
measured
in respiration
chamber
over
at least 20 min.
*P
energy expenditure measured by two independent in pregnant and nonpregnant Gambian women
A. HEINI, Y. SCHUTZ, E. DIAZ, A. M. PRENTICE, R. G. WHITEHEAD, AND E. JRQUIER Institute of Physiology, Faculty of Medicine, University of Lausanne, CH-1005 Lausanne, Switzerland; and Medical Research Council, Dunn Nutrition Unit, Keneba, The Gambia; and Cambridge CBA IXJ, United Kingdom A., Y. SCHUTZ, E. DIAZ, A. M. PRENTICE,R. G. AND E. J~QUIER. Free-Ziuing energy expenditure measured by two independent techniques in pregnant and nonpregnant Gambian women. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E9-E17, 1991.-Free-living energy expenditure (EE) was assessed in 37 young pregnant Gambian women at the 12th (n = 11, 53.5 t 1.7 kg), 24th (n = 14, 54.7 t 2.1 kg), and 36th (n = 12, 65.0 t 2.6 kg) wk of pregnancy and was compared with nonpregnant nonlactating (NPNL) control women (n = 12, 50.3 t 1.6 kg). The following two methods were used to assess EE: 1) the heart rate (HR) method using individual regression lines (HR vs. EE) established at different activity levels in a respiration chamber and 2) the doubly labeled water (2H2180) method in a subgroup of 25 pregnant and 7 control women. With the HR method the EE during the agricultural rainy season was found to be 2,408 t 87, 2,293 t 122, and 2,782 t 130 kcal/day at 12,24, and 36 wk of gestation and were not significantly different from the control group (2,502 t 133 kcal/day). These findings were confirmed by the ‘H2180 measurements, which failed to show any effect of pregnancy on EE. Expressed per unit body weight, the free-living EE was found to be lower (P < 0.01 with ‘Hz”0 method) at 36 wk of gestation than in the NPNL group. It is concluded that, in these Gambian women, energy-sparing mechanisms that contribute to meet the additional energy stress of gestation are operating during pregnancy (e.g., diminished spontaneous physical activity). HEINI, WHITEHEAD,
energy metabolism; physical activity
heart rate; doubly labeled water; pregnancy;
PREGNANCYAND LACTATION are two physiologicalproc-
esses of anabolic nature that increase the energy requirement of the mother (13, 14). In pregnancy the energy requirements are increased for two reasons. First, there is a rise in energy expenditure (EE) due to an increase in basal metabolic rate and a higher cost of movernent by the pregnant woman who gains weight. Second, there is an enhanced energy storage both of fetal and maternal origins (14). It is, however, not known whether these energy requirements must be wholly met by an increased food intake or whether adaptive mechanisms can spare energy during pregnancy. The assessment of 24-h EE in free-living populations of Third World countries is of great importance for two main reasons. First, in its latest report, the FAO/WHO/ UNU (13) expert committee recommended that energy requirements of adults should be assessed on the basis 0193-1849/91
$1.50 Copyright
of EE rather than energy intake data. Second, women living in Third World countries not only appear to be more physically active than those in industralized countries, but they are also pregnant more frequently and lactate longer, despite a level of food intake that has only been shown to be chronically marginal (23) by international standards (12, 13). The purpose of the present study was to assess EE in free-living conditions in pregnant and nonpregnant women living in a rural village of The Gambia. Two independent methods were used to assess EE [heart rate (HR) and doubly labeled water (2H2’“o)] to find out whether adaptive mechanisms to the marginal energy intake (23) occur during pregnancy. MATERIALS AND METHODS
The study was carried out in women from three rural Gambian villages (Keneba, Manduar, and Kantonkunda) of the subsahelian region. A detailed description of the environmental conditions prevailing in these villages has been previously published (16). The study period covered the beginning of the rainy agricultural season (July, 1987) to the dry season (April, 1988). The rainy season is characterized by heavy farming work and food shortage, whereas, during the more favorable dry season, the food availability is increased. Thirty-seven young pregnant women and twelve matched nonpregnant nonlactating (NPNL) women, who served as controls, took part in this crosssectional study. The pregnant women were studied at 12 (n = ll), 24 (n = 14), and 36 (n = 12) wk of pregnancy. Pregnancy was excluded or confirmed by a pregnancy test. Assessments of gestational age at birth were based on Dubowitz scores (8) by retrospective calculations. The women were apparently healthy, and none had complications or presented any clinical signs of malnutrition. The physical characteristics of women are shown in Table 1. The women gave birth to babies whose weight averaged 3.05 t 0.45, 2.75 t 0.29, and 2.96 t 0.26 kg for the women studied in the first, second, and third trimesters of pregnancy, respectively. Acceptance to participate in the study was obtained after giving a detailed explanation of the procedure to be used. The study was approved by the Medical Research Council/Gambian Governmental Ethical Committee. Anthropometry and body composition. Body weights were measured with a platform digital scale that was
0 1991 the American
Physiological
Society
E9
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
El0
ENERGY
TABLE
1. Physical characteristics Subject
Age,
1 2 3 4 5 6 7 8 9 10 11 12 Means&SE
22 37 24 25 26 32 24 32 26 24 30 29
Yr
EXPENDITURE
IN
PREGNANT
GAMBIANS
of pregnant and nonpregnant no&acting women
Weight, kg
Height, cm
BMI, kg/m2
Height2
Body Fat, % wt
FFM, kg
29.9
20.0 28.7 18.7 16.5 22.6 17.2 21.5 29.5 19.7 18.9 26.3 22.5t1.4
39.1 24.1 37.6 34.6 36.4 40.9 40.1 43.8 35.9 36.6 47.0 40.6 38.9tl.l
45.5 36.4 38.4 40.5 38.3 41.1 43.1 57.3 36.1 35.7 49.6 41.0 41.021.8
21.9 24.8 19.5 25.4 26.5 24.4 20.2 20.3 28.7 21.2 19.4 22.9t1.0
38.8 39.9 39.5 39.7 45.1 42.8 39.3 37.1 46.1 43.4 39.9 41.1t0.9
43.1 46.6 41.0 48.0 43.1 40.3 35.7 49.4 49.3 40.0 43.7t1.5
18.8 23.5 23.0 24.5 32.3 26.5 19.5 20.9 22.1 23.1 20.4 30.3 22.8
39.0 44.4 50.4 37.6 39.5 42.4 42.3 38.2 42.6 41.5 41.1 35.7 49.8 38.3 41.6t1.5
39.3 46.0 52.0 35.2 44.8 43.7 51.1 43.6 49.3 45.0 46.4 37.7 44.7 43.9 44.5k1.3
Body
Resistance,
cm2/C!
NPNL
27.6t1.3
55.8 42.6 52.8 42.6 43.6 52.8 48.4 55.8 51.0 45.6 58.0 55.1 50.3t1.6
153.0 156.0 162.5 150.0 163.0 155.0 161.0 167.0 153.5 154.0 168.0 160.0 158.5t1.5
49.7 53.0 49.1 53.2 61.3 56.6 49.3 46.6 64.7 55.0 49.5 53.5t1.7
155.0 155.5 158.0 159.0 166.0 167.5 165.5 152.5 165.0 155.5 161.0 160.1t1.6 156.0 160.0 172.0 168.0 162.5 159.5 160.5 154.0 163.0 159.0 157.5 151.5 157.0 156.0 159.8t1.4 155.0 164.0 163.0 179.0 164.0 155.5 160.5 163.0 169.0 161.5 167.0 156.5 163.1t1.9
23.8 17.5 20.0 18.9 16.4 21.9 18.7 20.0 21.5 19.2 20.5 21.5 20.0t0.6 1 st Trimester
13 14 15 16 17 18 19 20 21 22 23 Means&SE
20 28 22 23
18 25 25 30 32
21 26
24.6t1.3
20.7 21.9 19.7 21.0 22.2 20.2 18.0 20.0 23.8 22.7 19.1 20.9t0.5 2nd Trimester
24 25 26 27 28 29 30 31 32 33 34 35 36 37 MeanstSE
28 25 32 17 35 26 22 35 22 21 28 22 26.1t1.4
48.0 58.0 65.5 47.2 52.3 62.6 57.6 47.5 53.9 53.3 53.4 44.9 71.4 49.6 54.7t2.1
38 39 40 41 42 43 44 45 46 47 48 49 MeansHE
20 19 31 23 21 20 28 34 22 23 29 33 25.3t1.6
58.0 66.6 64.3 81.1 62.3 57.1 60.7 60.1 72.1 80.0 66.0 51.9 65.0&2.6*
25 27
19.7 22.7 22.1 16.7 19.8 24.6 22.4 20.0 20.3 21.1 21.5 19.6 29.0 20.4
21.4t0.8
23.4tl.l
3rd Trimester
NPNL,
nonpregnant
nonlactating;
BMI,
body
mass index;
FFM,
24.1 24.8 24.2 25.3 23.2 23.6 23.6 22.6 25.2 30.7 23.7 21.1 24.420.7t fat-free
accurate within 50 g. The subjects were under fasting conditions and were wearing a typical African cloth, the weight of which (-0.3 kg) was subtracted from the total body weight. Height was measured with a stadiometer, the accuracy of which was within 0.5 cm. Body composition was assessed using the following
22.3 45.1 46.0 31.7 45.5 50.6 25.5 47.9 46.9 25.8 60.2 57.4 23.6 47.6 52.3 20.6 45.3 40.8 19.8 48.7 45.8 21.0 47.5 54.3 26.7 52.9 58.3 32.2 54.2 64.7 21.8 51.6 51.4 25.0 38.9 46.8 24.7tl.l 48.8k1.61 51.3+1.9? mass. * P < 0.01, t P < 0.05, and $ P < 0.001 vs. other groups.
two independent methods: 1) skinfold thicknesses measured at four sites (biceps, triceps, subscapular, and suprailiac) (10) and 2) whole body resistance (20) measured at frequencies of 0.5, 50, and 100 kHz with a portable device developed at the Institute of Physiology (unpublished observations). For each subject, height squared
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
ENERGY
EXPENDITURE
IN
GAMBIANS
El1
all periods in the calorimeter during which the daytime HR was below this critical HR rate, an average value of sedentary EE was calculated for each subject (Fig. 1). This average value was used in everyday life situations to compute EE when HR was below the critical value. Leg work predominated in the treadmill exercises that were used to establish the relationship between EE and HR in the respiration chamber. However, visual assessment of the work done by the women in the fields shows that physical activity involved both arm and leg work. To ascertain whether the regression line was influenced by the nature of the work performed, a different group of subjects (9 NPNL women aged between 19 and 39 yr) was studied; they mimicked the type of activities performed in the field while in the respiration chamber. The exercise consisted in moving a l-kg block of wood on a horizontal surface and lifting it to two different heights (15 and 30 cm) each, over a lo-min period. In addition to the exercise period, two sitting values were measured over 10 min before exercise. For each subject the linear regression lines obtained with this combination of arm plus leg work was compared with that previously obtained with the treadmill exercise. In free-living conditions, HR was recorded during normal daily activities. The investigator went to the subject’s home at 7:30 A.M., placed the electrode belt on the chest, checked HR, and initiated the recording. Similarly, in the evening, the investigator went to stop the recording and to collect the HR monitor at -7:30 P.M. The data were then immediately printed in the laboratory. In each subject, the duration of measurements ranged from 1 to 4 days. The sedentary rate of EE previously described was used to cover the period between 7:30 P.M. and the time when the subject went to bed, typically 2 h later. During the night, HR was not monitored; the nighttime sleeping EE measured in the chamber was used to predict that at home. Calculation of EE. The daytime free-living EE was assessed as follows: after monitoring the minute-by-minute HR during daytime a histogram of HR was generated
divided by whole body resistance at 50 kHz was calculated. (Table 1). Assessment of total EE in free-living conditions. Total daily EE was assessed by two independent methods, a modification of the HR method originally described by Bradfield (4) and the ‘H2180 method in a subgroup of pregnant (n = 25) and NPNL women (n = 7). HR method. This method relies on the relationship between HR and EE and the subsequent measurement of HR in free-living conditions. A respiration chamber was used to assess EE at different activity levels (5, 21). The respiration chamber measures 5.1 x 2.4 x 2.4 m and has a net volume of 27 m3. It was built according to the model previously designed in Lausanne, Switzerland (15). The chamber is maintained at a temperature of 26°C and mixing of the air inside is ensured by an air conditioning system recirculating 1,000 m3 air/h. The performance of the calorimeter was tested using infusions of mixtures of CO2 and nitrogen. The accuracy averaged 99% for O2 and 100% of the “true” value for CO2 with a precision of 4-10 ml/min over a l-h period. HR were recorded by means of one of the following two portable instruments: 1) the “Sport Tester” (Polar electro, Kempele, Finland) and 2) the “Baumann” HR counter (Fleurier, Neuchatel, Switzerland). Both devices are small and light. The HR data were averaged over 1-min intervals and were stored in a memory. To study the relationship between HR and EE, five treadmill exercises of various intensity [speeds 1.4, 1.6, and 2.0 miles/h (mph) on the flat or on slopes of 5 and lo%] were carried out. Each exercise lasted for 15 min. In addition, two sitting values were obtained and a linear least-square regression line between HR, and EE was calculated through the experimental points (Fig. 1). A critical HR was determined below which the relationship between EE and HR was not linear (28). For each individual this value was defined as the midpoint between the HR of the treadmill exercise at the lowest intensity (1.4 mph, flat) and the mean HR obtained during the two sitting periods. A HR below this critical value was considered a “sedentary” level of activity; for 4o .
PREGNANT
[kcal/min]
1 I
3.5 . 3.0 I 2.5 . 2.0 .
1.5 -
EEsed
. ~1,.,111111111111111~~..~~~...~.~ ,,,,,,,,,,,,11,*,,*
1.0 II 0.5 I 0.0 -
I
70
I
Sedentary activity
D
FIG. 1. Individual heart rate (HR) calibration in respiration chamber in 1 woman. HR,,it, critical HR; EE,,d, sedentary energy expenditure; Exe, sitting; Exl-Ex5, treadmill exercise.
: f I
80
3
I
i 90 HR crit
I
I
100
I
I
110
I
I
120 Heart rate [beatslmin]
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
El2
ENERGY
EXPENDITURE
I 54
62
70
78
86
94
102
110
118
126
Heart
5
0 LJ
54
62
70
86
94
102
rate
IJW
110
118
rate
[beets/mlnI
20
0 54
62
70
78
86
94
102
110
118
!26
Heart
2. Typical HR distribution and in free-living conditions
FIG.
(A)
in 1 subject over 2 different
rate
PREGNANT
GAMBIANS
fasting. Serial mid-morning urine samples were collected for the next E-14 days. Isotope enrichment levels were measured using a dual isotope-ratio mass spectrometer (VG Isogas Aquasira, Middlewich, Cheshire, UK). CO2 production was calculated using Coward’s modification (6) of the original Lifson formula (18). The fractionation correction factors used here were 0.99 for 2H20, 1.04 for C802, and 0.93 for H2180. The proportion of water fractionated was calculated for each individual as described elsewhere (24). CO2 production rates were converted to EE using classical indirect calorimetric equations (7a), which require a value for the mean respiratory quotient (RQ) over the measurement period. An assumed RQ of 0.90 derived on the basis of the high-carbohydrate diet consumed in this community was used (2). Propagation of error analysis, as described by Coward et al. (7), was performed on each measurement to obtain an estimate of individual errors. Data analysis. Statistical analyses were performed by using a one-way analysis of variance followed by the multiple comparison test of Tukey. All results presented are expressed as means -+ SE. RESULTS
126
Heart
C
[beats/mln]
IN
[beats/mln]
in respira tion chamber days (B and C).
(Fig. 2). By referring each individual HR to the regression line of HR vs. EE (obtained in the respiration chamber) the daytime free-living EE was calculated. The total EE was obtained by simply summing the three following periods: 1) the daytime EE, 2) sedentary EE before going to bed, and 3) nighttime sleeping EE. ‘&180 method. Total free-living EE was measured using the 2H2’80 method (18). Due to the high cost of isotope, we used this technique only in a subset of women. The principle of the technique together with the full details of experimental method, calculation of results, validation, and limits of error have been described elsewhere (19,24). The subjects received an oral dose of 0.05 g/kg ‘H20 and 0.15 g/kg H2180 after an overnight fast and after the collection of a predose baseline urine sample. Postdose urine samples were collected after a further 4-h eouilibration period during which the subiect was
Antropometry and body composition. The average body weights of the groups of pregnant women were 53.5,54.7, and 65 kg for the first, second, and third trimester groups, respectively, whereas the control group had an average body weight of 50.3 kg (Table 1). The average height of the women in each group was not significantly different from each other. As expected, the average body mass index of the third trimester group was significantly greater than that of the other groups (P < 0.05). In the third trimester the fat-free mass was 25% greater than in the NPNL group when assessed from skin folds (P < 0.001) and when estimated from specific whole body resistance (P < 0.05), respectively (Table 1). HR. A typical HR distribution of one woman obtained in the chamber and in free-living conditions is shown in Fig. 2. Mean daytime HR measured in free-living conditions, resting HR measured in the respiration chamber, and the ratios of these two values are shown in Table 2. Average daytime HR tended to be higher in the pregnant (99-103 beats/min) than in the NPNL women (97 beats/ min), but the difference was statistically significant only in the second trimester (P < 0.05). As expected, resting HR were also higher during pregnancy, but this difference reached statistical significance only between the third trimester (72 t 3 beats/min) and the NPNL groups (61 t 2 beats/min) (P < 0.01). The critical HR was found to be in the order of 84-89 beats/min during pregnancy and 79 beats/min in the NPNL group. The percent of the daytime spent below the critical HR value increased as pregnancy advanced, from 26% in the NPNL group to 33% during the third trimester (Table 2). Regression lines between HR and EE. The average regression characteristics and correlation coefficients of EE vs. HR are shown in Table 3 for each group of women. In the third trimester group. the slope of the regression
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on November 5, 2018. Copyright © 1991 American Physiological Society. All rights reserved.
ENERGY TABLE
2. Heart
IN
PREGNANT
El3
GAMBIANS
rates Mean
Subject
EXPENDITURE
Daily
HR
Critical HR, beats/min
n
Beats/min
1 2 3 4 5 6 7 8 9 10 11 12 Means&SE
2 2 2 2 2 1 2 2 2 2 2 2
97 97 90 94 90 95 110 96 97 99 103 96 97t2
75 86 88 68 68 69 80 88 83 75 84 87 79Ik2
13 14 15 16 17 18 19 20 21 22 23 MeanstSE
2 2 2 2 2 2 2 2 3 2 1
92 94 100 101 98 108 104 105 93 101 95 99t2
78 82 89 79 90 91 73 102 80 77 82 84t3
24 25 26 27 28 29 30 31 32 33 34 35 36 37 MeanskSE
2 3 2 2 2 2 2 2 1 2 1 2 2 2
100 93 94 103 110 105 108 97 109 109 107 103 109 101 103&Z*
96 92 85 94 98 85 81 70 91 78 80 100 93 86 88t2
38 39 40 41 42 43 44 45 46 47 48 49 MeanstSE
2 1 2 2 1 2 3 2 4 2 2 2
95 99 109 97 99 110 103 96 97 101 91 110 101&Z
89 99 86 70 85 99 94 85 87 104 71 94 89t3
%Time Critical
Below HR
Mean
Resting HR, beats/min
Daily
Resting
HR HR
NPNL 12 14 52 25 10 19 22 30 36 24 32 32 261k3
58 71 67 52 54 53 57 73 64 62 50 67 61k2
1.43 1.31 1.27 1.71 1.48 1.64 1.80 1.23 1.52 1.46 1.75 1.50 1.51t0.05
28 26 34 22 32 23 18 30 25 24 35 29t2
55 63 72 62 72 67 63 77 64 62 69 66t2
1.39 1.34 1.22 1.44 1.26 1.42 1.65 1.42 1.37 1.53 1.30 1.40t0.04
47 50 52 31 24 26 18 16 24 18 19 42 32 22 30t3
82 63 68 78 83 62 70 60 64 60 59 76 65 69 69k2
1.19 1.22 1.31 1.12 1.20 1.36 1.46 1.43 1.65 1.47 1.62 1.30 1.70 1.40 1.39t0.05
41 54 22 26 25 22 42 32 29 58 32 15 33t4
68 78 68 62 71 77 79 68 70 89 56 77 72+3t
1.32 1.21 1.42 1.39 1.34 1.38 1.39 1.33 1.35 1.20 1.60 1.38 1.36kO.03
1st Trimester
2nd Trimester
3rd Trimester
HR, heart rate; n, no. of days of measurements. 0.05 and t P < 0.01 vs. NPNL group.
Resting
HR
was the lowest
line was significantly steeper than in the second trimester group (P < 0.05). The slopes of the regression line of EE (kcal/min) on HR (beats/min) were essentially the same when a combination of arm and leg work (0.039 t 0.004) was compared with leg work only (0.039‘t 0.005).
HR
measured
in respiration
chamber
over
at least 20 min.
*P
