Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 Body Composition Changes During Pregnancy In the development of standards for optimum weight gain during pregnancy, or in the use of weight gain to identify suboptimal pregnancies, the variability in the components of weight gain must be recognized. These include the products of conception (fetus, placenta, and amniotic fluid), uterine and breast tissue, extracellular fluid, and maternal fat. These components change over the course of pregnancy and to different extents in different individuals, markedly affecting the interpretation of weight gain. Although measurement of weight gain can be a clinically useful screen- ing method for identifying some pregnancies that are progressing abnor- mally, it provides very limited information regarding changes in body com- position of an individual pregnant woman, even when weight gain is close to the average for normal pregnancies. Information on body composition would add substantially to understanding of the meaning of a given weight gain. Fetal growth may be influenced more by specific maternal tissue changes, for example, by accretion of lean tissue, fat, or body water, than by total gestational weight gain. Body composition studies in appropri- ate animals could provide valuable information in this regard. However, even if changes in lean tissue should prove to be more important for fe- tal outcomes, methods would still be needed to determine accurately the net amount of \fat stored during normal pregnancy for estimating energy requirements, since fat is the most calorie-dense substance deposited. 121
122 NUTRITIONAL STAN-US AND WEIGHT GAIN STANDARI) METHODS In the most widely used model for examining body composition, the body is regarded as being composed of only two compartments fat and lean. In this usage, lean body mass represents a mixture of all the nonfat tissues of the body. Most techniques currently used to estimate body composition are based on measuring the qualities of the lean body tissues. Of the commonly used methods, only density measurements are dependent on both fat and lean tissue, but the fat estimate is still highly influenced by the variability of the lean tissue density. In the two compartment model, the weight of fat is the difference between two large masses" body weight and lean body mass. Therefore, a small relative error in the lean body mass estimation will produce a much larger relative error in calculated body fat. There are three standard methods for estimating lean body mass: measurement of total body water, determination of total body potassium content, and underwater weighing, which permits estimation of total body density, thereby allowing simultaneous estimation of both fat and lean tissue. Inherent in each of these methods are assumptions relating the actual measurements to specific body compartments. The assumptions are discussed here to assist in later interpretation of the data. Total Body Water 1b calculate lean tissue from total body water, the water content of the lean tissue must be known. Although the average percentage of water in lean tissue is known with fair accuracy in adult women, the nonfat tissues added during pregnancy (edema fluid, fetus, amniotic fluid, plasma) contain a high percentage of water. Thus, pregnancy may increase the water content of lean tissue from approximately 72.5% at 10 weeks of gestation to about 75.0 at 40 weeks in women with generalized edema (van Raaij et al., 1988~. A difference of this magnitude can cause fat to be underestimated by 50% or more in women gaining 3 to 4 kg of fat. Since gestational changes in lean tissue hydration in individual women have not been measured in body water studies, only approximate correc- tions are possible. Theoretical corrections for dilution of the lean tissues during pregnancy may improve estimates of body composition changes for a population; lean tissue estimates for an individual (which are impor- tant for relating body composition to pregnancy outcome) may still be inexact, although they are useful for identifying markedly aberrant cases. Interpretation of body water changes might be improved with a measure of extracellular water. Variation in extracellular water can be substantial. Hytten (1980) estimated that pregnant women with generalized edema have more than 3 kg (6.6 lb) of additional extracellular fluid compared with that
BODY COMPOSITION CHANGES DURING PREGNANCY 123 in women with no edema or leg edema only. Extracellular water can be determined either with the use of an extracellular tracer such as bromide or by estimation of intracellular water from measurement of total body potassium and determination of extracellular water by difference from total body water. There have been few studies in which extracellular water has been measured with tracers appropriate for use in pregnant women. Three small studies (Emerson et al., 1975; Forsum et al., 1988; Pipe et al., 1979) combined total body water and total body potassium measurements. These are discussed below. Underwater Weighing Underwater weighing is based on the assumption that the weight of fat and lean tissue can be estimated from total body density by using standard values for the average densities of fat and lean tissues. Because of the increased hydration of lean tissue during pregnancy, and especially because added tissue includes little bone, which is dense, the density of the lean body mass is likely to decline during pregnancy. Using theoretical estimates of body composition during pregnancy, Fidanza (1987) estimated that the density of the fat-free body declines from 1.100 kg/m3 at 10 weeks of gestation to 1.087 at 40 weeks of gestation. If nonpregnancy lean tissue density values are used for pregnant subjects, lean body mass will be underestimated and fat will be overestimated, perhaps by as much as 2.5 kg at term (see van Raaij et al., 1988~. In addition, the true mean density of the lean body tissue differs among individuals because of differing proportions of the organs, muscle, and bone comprising the lean body and may also change to varying degrees over the course of pregnancy because of the differential growth of various tissues, especially those with a high water content and no bone. Total Body Potassium Measurement of total body potassium can be used to estimate lean body mass if a standard value for the concentration of potassium in the lean tissues is assumed. The vast majority of the body's potassium (approx- imately 98%) is intracellular; therefore, total body potassium is actually a reflection primarily of the intracellular compartment. Substantial changes in the extracellular compartment can go undetected. For this method to give a good estimate of the weight of total lean tissue, the ratio of intra- cellular to extracellular tissue must be either close to the norm or assessed independently. This ratio of intra- to extracellular water decreases during pregnancy, resulting in overestimation of fat if not corrected. Independent
124 NUTRITIONAL STATUS AND WEIGHT GAIN measures of hydration (total body water, extracellular water) allow correc- tion for individual variation, as has been done in some studies (Forsum et al., 1988; Pipe et al., 1979~. Despite their limitations, total body water, underwater weighing, and total body potassium are the three best methods for studying body compo- sition in pregnant women. However, measurement of total body potassium and underwater weighing require special, large equipment and considerable patient cooperation, and estimation of total body water requires special iso- topes that are expensive to use and measure. These considerations have encouraged the use of simpler methods, such as measurement of skinfold thicknesses with calipers. SKINFOLD THICKNESS MEASUREMENT Changes in skinfold thickness have been widely used to estimate changes in the fat content of pregnant women. Skinfold thickness mea- surements suggest that more maternal fat is accumulated centrally than peripherally (Taggart et al., 1967~. Skinfold thickness can be measured quickly with relatively inexpensive equipment. As one early researcher cautioned, however, ". . . skinfold measurements are relatively inaccurate and . . . a high degree of standardization is required to obtain reliable comparisons, even with one observer" Braggart et al., 1967, p. 441~. Proper use requires extensive training and monitoring to consistently achieve re- producible measurements. 1b convert skinfold thickness measurements to estimates of body fat, standard regression equations are used. Generally, these are based on studies correlating skinfold thickness to body fat measured by total body water, body density, or total body potassium. Of special importance is the fact that the most widely used regression equations for interpreting skinfold thicknesses in pregnant women (Durnin and Womersley, 1974) have been developed in studies of nonpregnant subjects. Longitudinal studies of skinfold thickness in pregnant women (Taggart et al., 1967) suggest that skinfold thickness in late pregnancy may be increased by water retention. Therefore, an observed increase in skinfold thickness may not indicate an increase in body fat. The magnitude of this hydration effect may also vary from one measurement site to another, as indicated by a decrease in some skinfold thicknesses between the final weeks of pregnancy and the first month post partum. This has been studied by Adair et al. (1984) in Taiwan, by Taggart et al. (1967) in Scotland, and by Forsum et al. (1989) in Sweden. Thus, especially during late pregnancy, skinfold thickness measurements may be less indicative of body fat content. Because skinfold measurements are used (despite their limitations) in many clinical settings, it would be of
BODY COMPOSITION C~4NGES DURING PREGNANCY 125 great value to develop calibration equations derived from pregnant women whose body fat was estimated with the best methods and models available. The applicability of the equations may also be affected by differences in age, ethnic background, and exercise patterns of the reference and study populations. The usefulness of a regression equation depends in part on the comparability of the measurement techniques used in the population under study and in the reference population from which the equation was derived. Different or less experienced workers in the same research group may obtain different values or measurement variabilities (braggart et al., 1967~. Therefore, even very exact regression equations obtained by one group of investigators may give less accurate estimates of body fat when applied in a new study. DIRECT COMPARISONS OF SKINFOLD THICKNESS MEASUREMENT S In several studies, skinfold thickness values themselves have been used without calculating body fat content. In this approach, investigators used either individual skinfold thicknesses (Frisancho et al., 1977; Maso et al., 1988; Viegas et al., 1987) or a sum of several different skinfold thicknesses (Arroyo et al., 1978; Lawrence et al., 1984; Prentice et al., 1981; Taggart et al., 1967), but they did not assume that all the measured change reflects changes in body fat. This approach may be conceptually more justifiable than relating skinfold thicknesses to body fat. Furthermore, by combining skinfold thickness measurements with arm circumference measurements, it is possible to estimate arm muscle area, which reflects the amount of lean tissue. This could be of value, since it is not known whether maternal fat or lean tissue increments are more important for fetal growth. Frisancho et al. (1977) observed that maternal arm fat in Peruvian women was related to infant fatness but not birth weight, whereas arm muscle area was related to infant length. In contrast, Maso et al. (1988) observed that arm fat area and arm circumference changes between weeks 22 and 32 of gestation were correlated with birth weight in a U.S. black population, but arm muscle area was not. Changes in total body water and plasma volume also reflect com- ponents of the lean tissue. Both these measurements have been related to birth weight (Duffus et al., 1971~. Thus, although fat changes may contribute most to gestational calorie needs, lean tissue may influence important aspects of fetal growth. Clearly, better understanding of the importance of lean tissue changes during pregnancy is needed. Develop- ment of this understanding will require further study of total body water, total body potassium, extracellular water, and plasma volume and their relationship to pregnancy outcome.
126 NUTRITIONAL STATUS AND WEIGHT GAIN TABLE 6-1 Changes in Total Body Water from Studies Covering Different Periods of Gestation Change in Country and Number of Total Body Period of Comments on Study Reference Subjects Water, liters Gestation, wk Subjects Scotland 82-91 7.74 + 2.63 (SD)a 10 to 38 Mixed parities Hytten et al., 84 1.88 10 to 20 1966 82 2.80 20 to 30 91 3.03 30 to 38 Scotland 48 7.3 10 to 38 Estimated from their Taggart et reported dry al., 1967b weight Scotland 35 3.2 3~32 to 3~39 Primiparous, under Duffus et al., age 30 1971 United States 5 4.4 20 to 40 Four subjects (Boston) restricted their Emerson food intake; one et al., 1975 was obese England 27 7.2 1(}14 to 3~38 Normal Pipe et al., 3.3 1~14 to 2~28 prepregnancy 1979 3.9 2~28 to 3~38 weight for height Scotland 81 3.8 30 to 38 Obese women Campbell, 1983 Sweden 22 5.7C Prepregnancy Twenty-one Forsum to 36 weeks multiparous et al., 1988 4.2 1~18 to 30 2.0 30to36 a SD = Standard deviation. b It is not clear whether the patients in the study by Taggart et al. (1967) are a subset of those in the study by Hytten et al. (1966~. Hytten's group had 39 primiparas and 54 multiparas and Taggart's group had 23 and 25, respectively, drawn from the same research site, at about the same time, with the same coauthors. c All other studies used deuterium oxide dilution to determine total body water. This study used water labeled with Oxygen. Several newer methods (e.g., total body electrical conductivity, bioim- pedance analysis, and computerized axial tomography) for measuring fat or lean tissue may produce accurate results quickly and relatively easily. Some have gained acceptance by being validated against more familiar methods. However, there are no formal reports of their application to pregnant women for consideration by this subcommittee.
BODY COMPOSITION CHANGES DURING PREGNANCY TABLE 6-2 Estimated Total Maternal Weight Gain, Corrected to 40 Weeks of Gestation,a from Six Studies 127 Total Weight Reported Mean Gain, kg Weight Gain, Period of (corrected to 40 Reference kg + SD Gestation, wk weeks of gestation) Hytten et al., 1966 11.15 + 3.34 1~38 13.0 Taggart et al., 1967b 11.0C 1~38 12.8 Emerson et al., 1975 9.2c NRd 9.2e Pipe et al., 1979 10.4C 1~14 to 3~38 12.0 Campbell, 1983 9.2c 2~38 13.8 Forsum et al., 1988 13.6 + 3.0 Prepregnancy 13.6 to delivery a See text for method used to correct to 40 weeks of gestation. b See footnote b in Table 6-1. c Standard deviations (SD) were not reported. d NR = Not reported. e Not adjusted. Mean of "maximum gain" as reported by the authors, based on reported prepregnancy weights. Some of the five subjects were limiting their food intake. RESULTS OF STUDIES Total Body Water Table ~1 presents data from studies of total body water during preg- nancy. The tabulated results illustrate the interpretive problems: wide variations in weight gains among studies (see Table 6-2), different periods used to compute the gestational increment in body water, and small sample sizes. A better comparison of the tabulated studies, which cover different gestational periods, can be made by normalizing all the weight gain figures to 40 weeks of gestation (Table 6-2~. A graph of maternal weight gain (Hytten and Leitch, 1971) was used to estimate the weight that would have been gained before and after the measurement periods listed in Able 6-2. On the basis of this standard, expected additional weight gains are 0.5 kg (1.1 lb) before 5 weeks, 1 kg (2.2 lb) before 10 weeks, 4 kg (8.8 lb) before 20 weeks, 1.3 kg (2.9 lb) after 37 weeks, and 0.8 kg (1.8 lb) after 38 weeks of gestation. Able 6-2 shows the estimated total weight gains for six studies. These numbers suggest a consistency in average weight gain among the studies, whereas the values given for the various periods actually measured did not. The data provided by Pipe et al. (1979) and Hytten et al. (1966), which include measurements at three and four gestational periods, respectively,
128 NUTRITIONAL STATUS AND WEIGHT GAIN TABLE 6-3 Estimations of Fat Gain from Body Water Studies Change in Body Equivalent Change in Number of Water, in Lean Body Fat Gain Subjects kg + SDa b Mass,C kg Weight, kg Estimate, kg Hytten et al., 75 7.74 + 2.63 8.5 ll.~:Sa 2.65 1966 Taggart et al., 48 7.3d 8.0 ll.Oa 3.0 1967 Emerson et 5 6.3d (2.0 kg 6.9 8.2e 1.3 al., 1975 added for 1~20 wk) Pipe et al., 27 7.2d (1(}12 to 3~ 7.9 10.4 2.5 1979 38 wk) Campbell, 81 8 5d (4 7 kg 9.3 12.8e 3.5 1983 added for 1~30 wk) Forsum et al., 22 5.7d,/ 6.3 11.7 1988 NOTE: All data are normalized to a period ranging from 10 to 38 weeks or more of gestation and are corrected for the extra hydration of pregnancy. a From 10 to 38 weeks of gestation. b SD = Standard deviation. C Assuming 91.1% water in added lean tissues at term; see van Raaij et al. (1988~. This corrects for the added hydration of pregnancy, which would otherwise result in an under estimation of fat. d Standard deviation not reported or not applicable. e One kilogram was subtracted from the estimate of weight gain for the entire pregnancy to eliminate gains before 10 weeks of gestation. f The values for body water and pregnancy weight gain are the reported values from prepregnancy to 36 weeks of gestation. provide a means of evaluating the body water values obtained in the other studies. Both data sets show a larger increment of body water later in pregnancy than early in pregnancy. In general, other studies that provide fewer longitudinal data are in agreement (Campbell, 1983; Duffus et al., 1971~. The early increments obtained by Pipe et al. (1979) or Hytten et al. (1966) can be used in combination with data from studies that only provide measurements for the last half of gestation to estimate total increases in body water and body fat. Table 6-3 presents the results of six studies, along with estimations where needed. The highest estimate for fat gain was obtained by Forsum et al. (1988~. Their result was influenced by the value they found for body water incre- ments, which was much lower than those reported by Hytten et al. (1966) and Campbell (1983~. Of these three studies, only the one reported by Forsum and colleagues involved the use of water labeled with the isotope Oxygen, rather than deuterium, and saliva, rather than blood, to moni- tor dilution of the isotonically labeled water. Unless a correction is made,
BODY COMPOSITION CHANGES DURING PREGNANCY 129 the use of deuterium oxide will lead to a higher estimate of body water compared with that from the use of Oxygen water (which is considered the more accurate method), because deuterium exchanges with nonaqueous hydrogen to a small extent. These methodologic differences may contribute to the lower body water values obtained by Forsum and colleagues. How- ever, since deuterium oxide values are corrected for hydrogen exchange, it is unlikely that this factor completely explains the reported differences. Nevertheless, the study by Forsum and colleagues is provocative, be- cause it is the first longitudinal body water study that includes actual prepregnancy measurements. In addition, both the body water and total body potassium values they obtained suggest a loss of lean tissue in early pregnancy, as discussed below. Other considerations suggest that caution be exercised in accepting the data of Forsum and colleagues. For example, Clapp et al. (1988) showed an increase of both fat and lean tissue during the first 7 weeks of gestation and from weeks 7 to 15 of gestation (based on skinfold thicknesses). I5ggart et al. (1967), however, reported no increase in skinfold thicknesses in a group of women followed from before concep- tion through early pregnancy. The finding by Forsum et al. (1988) that total body water at 6 months pOSt partum is lower than the prepregnancy value, while body fat is 3.2 kg (7 lb) above prepregnangy levels, is surprising. It is difficult to accept the fact that women were retaining so much fat while they lost enough weight to place them below their prepregnancy weight. van Raaij et al. (1988) found that a 1.7-kg (3.7 lb) increase in weight from the prepregnancy to the postpartum period was associated with a 1.5-kg (3.3 lb) increase in fat, as determined by densitometry. Further studies will be needed to provide certainty about the changes that occur during early pregnancy. Body Density (by Underwater Weighing) Body density has been measured in two studies of pregnant women. In the most recent study (van Raaij et al., 1988), a new approach was used to interpret the measurements, correcting for changing density as pregnancy advances. This approach can also be applied to the earlier study by Seitchik et al. (1963), who reported individual values. When consistent methods of calculation are used, the two studies on body density are in excellent agreement (Bible 6-4~. All the differences in fat gain could be due to the differences in weight gains and gestation periods studied. These two studies give corrected fat estimates that are within the range of values obtained from total body water measurements. Total Body Potassium Results from studies of total body potassium in adult pregnant women are shown in Figure 6-1. Comparison of the three data sets reveals large differences in the absolute values obtained for total body potassium, espe
130 C) so Ct A: ._ 04 Ct Cal ._ Ct o m Ct Em A: ._ Ct Ct ._ U: 0 to ~ ~4 C': ·- a: to Cal ~ ~ O O Ce Em m ~ ._ Z U) 04 of m° X ~4 ;` 0 0 m 3 o o o ~ ·- c~: o ~ . Z Cal x Do . - ._, ._ Ct Ct ;> ~ ~ oo o o o . . . - Ct ._ ._ ~7 _` 3 US l _' a_ lo: Do l 3 1 - Do 1 ~o ~ - o o ~ o o ~ ~ u~ ~ oO ~cr~ ~ ~ ~ 0 . . . .. . . . . cr ~ ~oo 0 N ~;;~\ v: c ·~ ct - ~ c: t4 · - t: oc ·s~ c~ c~ ~c~ - . - 2 ~ Z C40 ~o C) ~ C) ~ 5 e,.D Ct .= 04 ~: ~ 3 ~ 2 5 ~ ._, ~Ct Ce ,.~: .o 3 ~_ oo oo - s: C~ Ct ;^ ._ O ~ ~ O O O D O O O O O O ~ ~ ~ ~ ~4 ~V ~C c5 ~ ~ ca ~ C~) _ Ct ~ O^ o ~ ~0 _C ~ ~ -~ ~ C O o~ D ~ G.) c' ce ~ 3 3 3 --,_ ~ o ;> O ~ ~ ~L~ z ~ q5 o ~c o o ~ r~ ~D O ~00 ~ ~ 00 ~n - ~ c~ ~ ~ ~ o ~ 3 ~ I ~ I ~ 3 C~ C~ Ct ._, ._ Ct - ~L ~n ._ Ct - o U) - 3 C~ ~S ._ C~ s~ ;^ Ct o o C~ 3 2 C~ ~: o - C~ ~s 5 3
BODY COMPOSITION CHANGES DURING PREGNANCY 3.4 U. o E ._ ._ E in _ ._ In ~ - 0 0 o En Is o 2.2 Weeks After Conception 131 Emerson et al., 1973 (N = 5) ---I-- Pipe et al, 1979 (N = 27) - -is- Forsum et al., 1988 (N = 22) / - ~ ' 1 1 0 10 20 30 40 50 Postpartum FIGURE 6-1 Changes in total body potassium during pregnancy based on data from three studies. cially in late pregnancy-differences that do not correspond to the reported differences in body weight. For example, weight gain was lowest in the report of Emerson et al. (1975), but total body potassium changes were the highest. Comparison of the incremental gestational changes in total body potas- sium also reveals substantial differences among the studies. The incremen- tal value of Forsum and colleagues (1988) from early to late pregnancy is approximately one-third that of the other two studies. The total body potassium data provided by Forsum's group, if correct, also suggest the loss of lean tissue during early pregnancy. In fact, the reported increment from prepregnancy to term is so low that it is insufficient to account for the amount of potassium others have estimated to be required for the contri- bution of the conceptus alone. However, all three studies show postpartum figures that are very close to their early pregnancy figures. If one ignores the prepregnancy figures of Forsum and colleagues, the pattern of the data is more consistent with those from the other studies. The total body potassium value and an estimate of the conceptus contribution to it (i.e., 169 mmol of potassium at 36 to 38 weeks of gestation; Pipe et al., 1979) can be used to calculate the amount of potassium gained
132 NUTRITIONAL STATUS AND WEIGHT GAIN by the mother herself. In turn, this value can be used to estimate the change in maternal lean tissue in the three studies, assuming that the incremental lean tissue has 92 mmol of potassium per kilogram (Pipe et al., 1979~. The results indicate a maternal gain of 1.2 to 2.1 kg (2.6 to 4.6 lb) of lean tissue between early and late pregnancy, excluding the conceptus. This would include blood, mammary gland, and uterine increments. Subtracting this estimate of the mother's lean tissue gain and an estimate of the weight of the conceptus from the weight gained during the corresponding period provides an estimate of the increment in maternal stores, largely fat, gained by the mother. This method of estimation indicated a loss of 1.6 kg (3.5 lb) of fat (weeks 20 to 40 of gestation) in the study of five women who were restricting food intake (Emerson et al., 1975~; these women may have gained weight earlier. When data from the other two studies and total body potassium changes are used, the estimated maternal stores are 3.6 kg (7.9 lb) (Forsum et al., 1988) or 4.6 kg (10.1 lb) (Pipe et al., 1979~. Both groups of workers have combined their data on total body water and their data on total body potassium to calculate a corrected value for total body fat. In this approach, changes in hydration are taken into consideration in computing lean tissue from the total body potassium. Their corrected fat estimates were 1.6 kg (3.5 lb) (Forsum et al., 1988) and 1.87 kg (4.1 lb) (Pipe et al., 1979) from early to late pregnancy, which are quite different from the uncorrected values calculated from the total body potassium value only. The values for fat increments in women on unrestricted diets estimated from the three standard methods suggest very different calorie requirements for fat storage-i.e., nearly a 30,000-kcal difference between the low and high values. This represents a substantial portion of the estimated energy requirement for pregnancy and indicates the need for a better definition of the changes that occur in calorie requirements and energr partitioning in successful pregnancies. Skinfold Thicknesses Skinfold thicknesses have been used to describe normal body fat changes throughout gestation, to determine whether skinfold thickness is associated with fetal outcome or with supplementation in undernourished women, to identify women with unusually small or large changes in body fat during pregnancy, and to estimate the initial body fat content. Measured mean triceps values range from a low of approximately 10 mm (at term in ~iwanese women; Adair et al., 1984) to a high of 18.9 mm (at 22 weeks of gestation, in black teenagers having appropriate-for-gestational- age newborns; Maso et al., 1988~. Mean values for the sum of triceps, biceps, subscapular, and suprailiac skinfold thicknesses ranged from a low of 31.3 mm (4 to 6 weeks post partum; Gambian data reported by Durnin,
BODY COMPOSITION CHANGES DURING PREGNANCY 133 TABLE 6-5 Estimations of Total Body Fat and Increase in Fat During Gestation Based on Skinfold Thickness Measurements and the Equation of Durnin and Womersley (1974) Study and Country Estimated Total Body Fat in First Trimester, kg + SDa Fat Increment, kg Pipe et al., 1979 15.4 + 2.9 2.8b England Dibblee and Graham, 1983 14.1 + 3.9 4.4b England Langhoff-Roos et al., 1987 Not available 4.0C Sweden Durnin, 1987 Scotland 15.1 + 4.6 2.3d The Netherlands 17.7 + 4.9 2.0d The Gambia 10.3 + 2.5 0.6d Thailand 11.3 + 2.8 1.4d Philippines 11.2 + 3.4 1.3d a SD = Standard deviation. b Fat gain from first to last trimester. c Fat gain from weeks 17 to 37 of gestation. d Fat gain from week 10 of gestation to 4 to 6 weeks postpartum. 1987) to a high of 64.8 mm (at 17 weeks of gestation in Swedish women; Langhoff-Roos et al., 1987~. Differences of this magnitude may partly resect methodologic differences. Table 6-5 shows the values obtained for body fat in the four studies in which the regression equation of Durnin and Womersley (1974) was used to compute body fat changes from skinfold thickness measurements. In light of the fact that the equation was derived from data on nonpregnant women, the general consistency of these findings with those from more complex methods is reassuring. The values for women from industrialized countries are in the same range as those found by the methods discussed previously, but there is nearly a twofold difference between the highest and lowest estimates of body fat content changes. The range is even wider if the values for women from developing countries are included. The data presented by Durnin (1987) are based on skinfold thicknesses measured 4 to 6 weeks post partum. This may partly explain why they are lower than the values from the other studies. For example, Dibblee and Graham (1983) estimated a 4.4 kg (9.7 lb) fat increment between the first and third trimesters, based on skinfold thickness changes. Yet, only 1.3 kg (2.9 lb) of that estimated fat gain was retained at 4 weeks post partum. Although some fat may be lost post partum, it is likely that the increase in body water contributes to the increase in skinfold thickness during pregnancy;
134 NUTRITIONAL STATUS AND WEIGHT GAIN the water loss post partum may contribute to the decrease in the skinfold thickness. A study by Clapp et al. (1988) has provided some information on changes in skinfold thickness that occur very early in pregnancy. Six skinfold thicknesses were measured serially in 20 women, starting before pregnancy. The data indicate that body weight increased by 2 kg (4.4 lb) and body fat increased by 1.54 kg (3.4 lb) between the prepregnancy measurement and the seventh week of pregnancy. Thus, this study supports the possibility that maternal fat may already be increased above prepregnancy levels by the time most studies of body composition during pregnancy are begun. If so, then when measurements begin after the first trimester, increments in total body fat may be underestimated. Alternatively, these findings may indicate that the relationship between skinfold thickness and total body fat is altered very early in pregnancy. This possibility cannot be evaluated until more studies using serial measurements of total body water, total body density, or total body potassium are done during the periconceptional period. SUMMARY Issues to consider when examining results of pregnancy body compo- sition studies include the following: . Each body composition method is based on underlying assumptions, and correction factors are needed to adjust for changes in the lean body during pregnancy. Without these corrections, total body water tends to underestimate total body fat and both underwater weighing and total body potassium tend to overestimate it. · In the future, multicompartment models of body composition need to be used in studies of larger numbers of pregnant women. Attention must also be given to differences in the gestational period studied, weight gain, initial weight, maternal age, ethnic background, and parity. . Skinfold thickness may be useful for research purposes, but the currently used reference equations may not permit calculation of actual total body fat changes. Because of its potential for clinical as well as research use, measurement of skinfold thickness needs to be standardized against reference methods in a large number of pregnant women. · For the development of dietary and weight gain recommendations, more information is needed on the relationship of weight gain to body fat gain in individual women. Studies of the effects of composition changes on other outcomes are also needed.
BODY COMPOSITION CHANGES DURING PREGNANCY REFERENCES 135 Adair, L~S., E. Pollitt, and W.H. Mueller. 1984. The Bacon Chow study eject of nutritional supplementation on maternal weight and skinfold thicknesses during pregnancy and lactation. Br. J. Nutr. 51:357-369. Arroyo, P., D. Garcia, C. Llerena, and S.E. Quiroz. 1978. Subcutaneous fat accumulation during pregnancy in a malnourished population. Br. J. Nutr. 40:485~89. Campbell, D.M. 1983. Dietary restriction in obesity and its eject on neonatal outcome. Pp. 243-250 in D.M. Campbell and M.D.G. Gillmer, eds. Nutrition in Pregnancy: Proceedings of the Tenth Study Group in the Royal College of Obstetricians and Gy- naecologists, September, 1982. The Royal College of Obstetricians and Gynaecologists, London. Clapp, J.F., III, BALD Seaward, R.H. Sleamaker, and J. Hiser. 1988. Maternal physiologic adaptations to early human pregnancy. Am. J. Obstet. Gynecol. 159:1456-1460. Dibblee, L., and T.E. Graham. 1983. A longitudinal study of changes in aerobic fitness, body composition, and energy intake in primigravid patients. Am. J. Obstet. Gynecol. 147:908-914. Duffus, G.M., I. MacGillivray, and KJ. Dennis. 1971. Ihe relationship between baby weight and changes in maternal weight, total body water, plasma volume, electrolytes and proteins, and urinary oestriol excretion. J. Obstet. Gynaecol. Br. Commonw. 78:97-104. Durnin, J.V.G.A. 1987. Energy requirements of pregnancy an integration of the longitudinal data from the five-country study. Lancet 2:1131-1133. Durnin, J.V.G.A., and J. Womersley. 1974. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 32:77-97. Emerson, K., Jr., E.L Poindexter, and M. Kothari. 1975. Changes in total body composition during normal and diabetic pregnancy. Obstet. Gynecol. 45:505-511. Fidanza, F. 1987. The density of fat-free body mass during pregnancy. Int. J. Vitam. Nutr. Res. 57:104. Forsum, E., A. Sadurskis, and J. Wager. 1988. Resting metabolic rate and body composition of healthy Swedish women during pregnancy. Am. J. Clin. Nutr. 47:942-947. Forsum, E., A. Sadurskis, and J. Wager. 1989. Estimation of body fat in healthy Swedish women during pregnancy and lactation. Am. J. Clin. Nutr. 50:465~73. Frisancho, AR., J.E. Klayman, and J. Matos. 1977. Influence of maternal nutritional status on prenatal growth in a Peruvian urban population. Am. J. Phys. Anthropol. 46:265-274. Hytten, F.E. 1980. Weight gain in pregnancy. Pp. 193-233 in F. Hytten and G. Chamberlain, eds. Clinical Physiology in Obstetrics. Blackwell Scientific Publications, Oxford. Hytten, F.E., and I. Leitch. 1971. The Physiology of Human Pregnancy, 2nd ed. Blackwell Scientific Publications, Oxford. 599 pp. Hytten, F.E., A.M. Thomson, and N. Taggart. 1966. Total body water in normal pregnancy. J. Obstet. Gynaecol. Br. Commonw. 73:553-561. Langhoff-Roos, J., G. Landmark, and M. Gebre-Medhin. 1987. Maternal fat stores and fat accretion during pregnancy in relation to infant birthweight. Br. J. Obstet. Gynaecol. 94:1170-1177. Lawrence, M., F. Lawrence, W.H. Lamb, and R.G. Whitehead. 1984. Maintenance energy cost of pregnancy in rural Gambian women and influence of dietary status. Lancet 2:363-365. Maso, M.J. , E.J. Gong, M. S. Jacobson, D.S. Bross, and F.P. Heald. 1988. Anthropometric predictors of low birth weight outcome in teenage pregnancy. J. Adol. Health Care 9:188-193. Pipe, N.G.J., T. Smith, D. Halliday, C.J. Edmonds, C. Williams, and T.M. Coltart. 1979. Changes in fat, fat-free mass and body water in human normal pregnancr. Br. J. Obstet. Gynaecol. 86:929-940. Prentice, AM., R.G. Whitehead, S.B. Roberts, and A.A. Paul. 1981. Long-term energy balance in child-bearing Gambian women. Am. J. Clin. Nutr. 34:2790-2799.
136 NUTRITIONAL STATUS AND WEIGHT GAIN Seitchik, J., C. Alper, and A. Szutka. 1963. Changes in body composition during pregnancy. Ann. N.Y. Acad. Sci. 110:821-829. Taggart, N.R., R.M. Holliday, W.Z. Billewicz, F.E. Hytten, and A.M. Thomson. 1967. Changes in skinfolds during pregnancr. Br. J. Nutr. 21:439-451. van Raaij, J.M.A., M.E.M. Peek, S.H. Vermaat-Miedema, C.M. Schonk, and J.G.AJ. Hautvast. 1988. New equations for estimating body fat mass in pregnancy from body density or total body water. Am. J. Clin. Nutr. 48:24-29. Viegas, O.A.G, T.J. Cole, and B.A. Wharton. 1987. Impaired fat deposition in pregnancy: an indicator for nutritional intervention. Am. J. Clin. Nutr. 45:2~28.