Results: Mothers who were lighter had babies who had lower birth weight, were shorter, had ... Descriptors: foetal growth restraint; maternal nutrition; birth weight.
European Journal of Clinical Nutrition (1997) 51, 134±138 ß 1997 Stockton Press. All rights reserved 0954±3007/97 $12.00
Relationship between maternal nutritional status and infant's weight and body proportions at birth M Thame, RJ Wilks, N McFarlane-Anderson, FI Bennett and TE Forrester Tropical Metabolism Research Unit, University of the West Indies, Mona, Kingston 7, Jamaica
Objectives: To examine maternal nutritional status and its relationship to infant weight and body proportions. Design: Retrospective study of births from January±December 1990. Setting: University Hospital of the West Indies, Jamaica. Subjects: Records for 2394 live, singleton births, between 200±305 d gestation. Main outcome measures: Birth weight, crown heel length, head circumference, ponderal index, head circumference:length ratio, placental weight, placental:birth weight ratio. Results: Mothers who were lighter had babies who had lower birth weight, were shorter, had smaller heads and had a higher HC:L ratio. Shorter and thinner women had babies who had lower birth weights, were shorter, had smaller heads and lighter placentas. Thinner women also had babies with a lower placental:birth weight ratio, and their BMI's were not linearly related to ponderal index and HC:L ratio. Women whose ®rst trimester Hb levels were < 9.5 g/dl had babies with the lowest birth weight, crown heel length, placental weight and ponderal index. These measurements increased as the Hb levels rose to 12.5 g/dl but then fell at Hb levels > 12.5 g/dl. In the second and third trimester Hb levels were negatively associated with birth weight, crown heel length, head circumference, placenta weight and ponderal index. Conclusions: The data support the hypothesis that poor maternal nutrition is associated with foetal growth restraint. Poor maternal nutrition as indicated by low weight, height, and BMI are associated with smaller, shorter babies with smaller heads. Haemoglobin levels > 12.5 g/dl in pregnancy are associated with lighter, shorter, thinner babies, with smaller heads. Descriptors: foetal growth restraint; maternal nutrition; birth weight
Introduction Perinatal morbidity and mortality are related to growth restraint in utero which presents as low weight at birth. The evidence that the low birth weight infant (BWT < 2500 g) is at risk at birth, perinatally and in early life, is well documented (Kramer, 1987; Naeye, 1979). So is the role of maternal nutritional status in in¯uencing infant size (Kramer, 1987; Naye, 1979). Post maternal nutrition is associated with constrained foetal growth and therefore lower birth weight. Restriction of energy and/or protein intake by mother during pregnancy is associated with small size at birth in rats, mice, sheep and pigs (Widdowson, 1971; Chow, 1964; Zeman, 1969; McCance, 1974; Levy and Jackson, 1993; Widdowson and McCance, 1963). In humans, children of mothers who are poorly nourished before and during pregnancy are smaller at birth (Barker, 1994). Mothers who were small at birth in turn give birth to small babies (Barker, 1994; Stein and Susser, 1975). Recently, retrospective studies by Barker and colleagues have shown that the risk of chronic cardiovascular disease in adulthood is inversely related to birth weight, even within the normal range (Barker, 1994). Speci®c relationships were shown between different patterns of foetal Correspondence: Dr T Forrester. Received 28 May 1996; revised 17 September 1996; accepted 27 September 1996
growth restraint and type of chronic cardiovascular disease in adulthood. Thus, adults who were small babies with reduced birth weight had increased blood pressure, thin babies (low ponderal indices) developed high blood pressure and non insulin dependent diabetes, while disproportionate babies (short in relation to head circumference) were shown to be at risk as adults for raised blood pressure, LDL cholesterol, and ®brinogen (Barker, 1992; Barker, 1993). The relationships between impaired foetal growth and risk factors for chronic cardiovascular diseases have also been demonstrated in childhood (Law et al, 1991). In recent studies of Jamaican children aged 7±14 y, birth weight was inversely related to blood pressure, and crown heel length at birth was inversely related to glycosolated haemoglobin levels and serum total cholesterol concentrations. Blood pressure in children was also inversely related to maternal haemoglobin concentration (Forrester et al, 1996). However, thin maternal triceps skinfolds during pregnancy was a more powerful predictor of blood pressure in childhood than maternal haemoglobin concentration (Godfrey et al, 1994). In situations of marginal nourishment there is some evidence that mothers adjust their metabolic demands, hence potentially sparing nutrients for the development of the foetus (Poppit et al, 1994). However, where there is chronic undernutrition, the limits of adaptation might be exceeded and foetal growth become impaired. Inadequate
Maternal nutritional status M Thame et al
protein and energy intake may be common among some poorer mothers in a developing country and could contribute to foetal growth restraint (Kramer, 1987; Thongprasert et al, 1987). Evidence from studies in rats indicate a negative relationship between maternal protein intake and blood pressure in their offspring (Langley and Jackson, 1994). The age adjusted prevalence of hypertension in Jamaica is approximately 23%, and, based on the hypotheses above, this high rate might in part related to impaired fetal growth (Miall et al, 1972; Cooper et al, 1996). This study was carried out to examine the relationship between maternal nutritional status and weight and body proportions at birth as they may have implications for the incidence of chronic cardiovascular diseases in later life.
Subjects and methods We used routinely kept antenatal records of weight, height and haemoglobin concentration from the University Hospital of the West Indies (UHWI) to obtain information on maternal nutritional status. Mothers attended clinic for booking between 8±10 weeks gestation, when their weight, height and haemoglobin levels were recorded. In addition, weight was measured in antenatal clinic at 4 week intervals up to 32 weeks gestation, then fortnightly to 36 weeks and weekly thereafter to term. In some mothers, haemoglobin was again measured at twenty-®ve to twenty eight, and at thirty-®ve weeks of gestation. Haemoglobin levels were measured on a Coulter Counter model T899 using standard automated techniques. Measurements of the newborn's birth weight, crown heel length, and head circumference are routinely made at the UHWI by midwives. Ponderal index, PI, (wt/ht3) and head circumference to length ratio HC/CHL, an index of disproportion, were calculated. 3054 patients delivered at the UHWI between January 1 and December 31, 1990. Records for 92% (2804) of these were available for inspection. Data are expressed as means s.d. Maternal weight, height, and BMI were divided into quintiles, and haemoglobin levels in each trimester were divided into categories. Minimum haemoglobin, the lowest haemoglobin recorded for a mother during pregnancy, was identi®ed and categorized similarly. Analysis of variance and tests for linear Table 1
trend were used to compare means. Multiple regression analysis was used to examine relationships between maternal characteristics and birth outcomes. Analysis was performed using the SPSS-PC statistical package.
Results Analysis was con®ned to 2394 infants. This represented 78% of all deliveries for the year 1990, and 85% of all the dockets available for inspection. The remaining 15% were omitted from the analysis, if women had had multiple births, or delivered foetuses with gestational age less than 200 d. Mothers' mean age and BMI were 24.6 y and 24.2 kg/ m2 respectively. Mean haemoglobin levels for each trimester were in the normal range (Table 1). Babies were 3.1 kg on average, but there was a wide range of birth weights in this population (Table 1). Mothers who were lighter had babies who had lower birth weight, were shorter, had smaller heads and had a higher HC/L ratio (Table 2). They had lighter placentas and lower placental:birth weight ratios. Maternal weight was not related to ponderal index (Table 2). Maternal weight gain in pregnancy was not related to any newborn measurement, but was weakly related (B 1.12 0.51, P 0.03) to placental weight in a multiple regression analysis. Shorter women had babies who had lower birth weights, were shorter, had smaller heads and smaller placentas. Maternal height was not related to ponderal index, HC/L ratio and the placental:birth weight ratio (Table 2). Thinner mothers had babies who had lower birth weights, were shorter, had smaller heads, lighter placentas and lower placental:birth weight ratios. BMI was not linearly related to ponderal index or HC/L ratio; however those women with the lowest BMI's had babies with the lowest PI and highest HC:L ratio (Table 2). Birth weight, crown heel length, placental weight and ponderal index were lowest in mothers with ®rst trimester haemoglobin levels 9.5 g/dl (Table 3). These measurements tended to increase as the haemoglobin levels rose to 12.5 g/dl, but then to fall off at haemoglobin levels > 12.5 g/dl. The test for linear trend was only signi®cant for ponderal index. In the second and the third trimesters, haemoglobin levels were negatively associated with birth
Maternal and newborn characteristics. Values are means s.d
Variable
Means s.d
Age (y) 1st trimester weight (kg) Maternal height (cm) BMI (kg/m2) Parity 1st trimester haemoglobin (g/d) 2nd trimester haemoglobin (g/dl) 3rd trimester haemoglobin (g/dl) Minimum haemoglobin (g/dl) Placental weight (g) Birth weight (g) Crown heel length (cm) Head circumference (cm) Gestational age (days) Ponderal Index (kg/m3) Head circumference : length ratio Placental : birth weight ratio
26.4 5.3 64.4 13.5 163.0 6.7 24.2 5.0 0.9 1.1 12.0 1.0 11.0 1.1 11.4 1.2 10.8 1.0 603 127.9 3190.6 527.3 52.61 4.0 34.4 1.8 273.2 14.3 22.32 4.4 65.7 4.5 19.07 3.3
BMI Body mass index. (Variability in n re¯ects the availability of information from the records).
Range
n
13±47 33.0±139.0 121.9±198.1 13.7 48.3 0±9 5.4±15.5 4.9±15.3 4.5±16.4 4.9±13.9 190±1300 670±4895 32.0±75.5 25.0±40.0 200±305 8.3±47.0 48.5±82.4 6±41.6
2385 2282 1559 1559 2388 2164 1879 1750 2108 2378 2381 2175 2176 2394 2174 2168 2365
135
Maternal nutritional status M Thame et al
136
Table 2 Relationship between maternal weight, height, BMI and birth outcomes Maternal weight (kg) 7 53 7 59 7 65 7 74 > 74 Maternal height 7 157 7 160 7 164 7 167 > 167 BMI 7 19 7 22 7 24 7 28 > 28 Maternal age (years) 7 21 7 24 7 27 7 30 > 30
BWT (g)
CHL (cm)
HC (cm)
HC : L ratio
PLWT (g)
PI
P/B ratio
n
3014 3118 3192 3310 3355 ***
51.6 52.5 52.6 53.3 53.3 ***
34.1 34.2 34.5 34.8 34.7 ***
66.3 65.4 65.8 65.4 65.5 *
561 585 596 632 646 ***
20.8 20.4 20.4 19.8 20.6
18.7 18.9 18.8 19.2 19.4 ***
464 421 454 457 477
3063 3119 3186 3179 3230 ***
52.2 52.2 52.4 52.7 52.9 **
34.1 34.2 34.3 34.4 34.6 ***
65.8 65.7 65.8 65.7 65.7
576 594 611 596 608 **
20.1 20.3 21.0 20.2 20.2
19.0 19.1 19.3 18.8 19.0
243 191 489 211 396
2969 3078 3188 3269 3311 ***
51.4 52.2 52.5 53.2 53.2 ***
33.9 34.2 34.3 34.5 34.7 ***
66.1 65.7 65.6 65.2 65.5
550 575 606 625 640 ***
19.9 20.7 20.8 20.1 20.2
18.6 18.8 19.2 19.2 19.4 ***
206 384 296 349 320
3089 3148 3210 3271 3220 ***
52.0 52.3 52.6 52.9 53.1 ***
34.2 34.4 34.4 34.6 34.6 ***
65.9 66.0 65.7 65.7 65.4 *
582 595 606 617 614 ***
22.2 22.3 22.6 22.6 21.9
19.0 19.0 19.1 19.0 19.2
447 448 525 443 510
ns
ns
ns
ns
ns
ns
ns
ns
BWT Birth weight. CHL Crown heel length. HC Head circumference. HC : L Head circumference : Length ratio ( 6 100). PLWT Placental weight. PI Ponderal index. P/B Placental : Birth weight ratio ( 6 100). n Number. * P < 0.05; ** P < 0.01; *** P < 0.001 (test for linear trend).
weight, crown heel length, head circumference, placental weight and ponderal index. The relationship with birth weight and third trimester haemoglobin were best descried by a quadratic function, with lowest birth weights observed at haemoglobin concentrations less than 9.5 g/dl and greater than 12.5 g/dl. There was no association with HC/ L ratio, and placental:birth weight ratio. Minimum haemoglobin, which was the lowest haemoglobin level at any time during pregnancy was also negatively associated with birth weight, crown heel length, head circumference, placental weight and ponderal index. These associations were all signi®cant except for ponderal index and head circumference:length ratio. Mothers whose minimum haemoglobin levels did not fall below 12.5 g/dl had babies who weighed less, were thinner, shorter, and had smaller heads. Their placentas were also smaller but there was no association with head circumference:length ratio (Table 3). The subgroup of women with haemoglobin in excess of 12.5 g/dl at booking were similar to the other groups of women in anthropometry and age. However, blood pressure was higher, systolic blood pressure 113 mmHg vs 110 mmHg, P < 0.001. The prevalence of hypertension in pregnancy and pre-eclampsia was not different in this subgroup compared with the others. Younger mothers had babies who were lighter, were shorter, had smaller heads, and lighter placentas. The head circumference:length ratio was highest in the younger mothers, but ponderal index did not vary (Table 2). However, in a multiple regression analysis, age did not make an independent contribution to birth weight after taking
account of mother's weight, height, parity, child's gender and gestational age. In order to identify the independent predictors of birth outcome, the variables with signi®cant Pearson correlation coef®cient were included in a regression analysis. The independent variables were gestational age, gender, maternal age, weight, height at booking, haemoglobin in each trimester, minimum haemoglobin during pregnancy and weight gain in pregnancy. With birth weight as the dependent variable 24% of the variance was explained by gender (girls were lighter), gestational age, maternal ®rst trimester weight and height. Neither maternal age nor parity made a signi®cant contribution. For length at birth (crown heel) 9% of the variance was explained by gender (girls shorter) gestational age, maternal weight and height at booking, and minimum haemoglobin in pregnancy (inverse). Likewise, 11% of the variance of head circumference was explained by gender (girls had smaller heads) gestational age, maternal ®rst trimester weight, and minimum haemoglobin. Weight gain in pregnancy made no independent contribution to outcome of any of the three newborn measurements. Discussion Some of the accepted indices of maternal nutritional status are weight, height, BMI, and Hb levels; our data show that babies who were born lighter, thinner, and disproportionate had mothers who were younger, lighter, shorter and thinner, and had higher Hb after the ®rst trimester during pregnancy.
Maternal nutritional status M Thame et al
137
Table 3 Relationship between haemoglobin in pregnancy and birth outcome Hb1 (g/dl)
BWT (g)
CHL (cm)
HC (cm)
HC : L ratio
PLWT (g)
PI (kg/m3)
P/B ratio
7 9.5 7 10.5 7 11.5 7 12.5 > 12.5 P Hb2 (g/dl) 7 9.5 7 10.5 7 11.5 7 12.5 > 12.5 P Hb3 (g/dl) 7 9.5 7 10.5 7 11.5 7 12.5 > 12.5 P Min Hb 7 9.5 7 10.5 7 11.5 7 12.5 > 12.5 P
2958 3227 3190 3198 3189
51.4 52.9 52.6 52.6 52.7
34.4 34.8 34.4 34.4 34.4
67.1 66.1 65.8 65.6 65.6
564 603 599 608 601
20.3 20.7 20.8 21.1 19.3 **
19.3 18.8 18.8 19.2 19.1
3226 3241 3201 3202 3128 **
52.9 52.9 52.7 52.6 52.2 **
34.8 34.5 34.5 34.1 34.2 ***
66.1 65.5 65.7 65.7 65.8
620 607 608 603 591 **
21.2 20.6 21.3 20.4 18.9 ***
19.3 18.8 19.1 19.0 19.2
3167 3218 3290 3199 3115 ***
53.0 52.8 53.3 53.6 52.1 ***
34.7 34.8 34.8 34.5 34.0 ***
65.9 66.2 65.5 65.8 65.7
600 615 611 602 596 *
20.0 20.5 20.8 21.0 19.8
19.3 19.2 18.7 18.9 19.4
3224 3217 3233 3142 3090 ***
52.9 52.9 52.7 52.2 52.2 **
34.8 34.6 34.5 34.2 34.0 ***
66.0 65.7 65.8 65.8 65.5
613 605 611 600 580 **
20.9 20.6 21.3 20.8 17.1 ***
19.2 18.9 19.0 19.3 19.2
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
BWT Birth weight. CHL Crown heel length. HC Head circumference. HC : L Head circumference : Length ratio (100). PLWT Placental weight. PI Ponderal index. P/B Placental : Birth weight ratio (100). Hb1 Haemoglobin in the ®rst trimester. Hb2 Haemoglobin in the second trimester. Hb3 Haemoglobin in the third trimester * P < 0.05; ** P < 0.01; *** P < 0.001 (test for linear trend).
Regression analysis indicated that maternal booking weight was the strongest predictor of birth outcome. The retrospective nature of the study did not allow us to obtain prepregnancy weights, so mother's weight at 8±10 weeks postconception was used as a proxy for her prepregnancy weight. Gueri et al (1982), have suggested that there is an increase of only 1.7% over the prepregnant weight during the ®rst 13 weeks of pregnancy. In this group of mothers, 13% had a BMI 19, suggesting that they entered pregnancy on a lower nutritional plane, and we have shown previously that mothers with thinner triceps skinfolds had children who at age 11 y had higher blood pressures (Godfrey et al, 1994). Our ®nding that women with lower BMI gave birth to babies who were lighter and shorter and had smaller heads are similar to those of several studies which have examined the relationship between maternal size and foetal outcome (Stein and Susser, 1975; Pivalizza et al, 1990; Baqui et al, 1994; Mavalankar et al, 1994). Pivalizza et al (1990), found that among South African women, those who produced light for date infants were the same height as study controls but they weighed less, suggesting that weight-forheight was an important determinant of reduced fetal growth. In a study in India, low maternal weight was found to be the major risk factor for low birth weight (Mavalankar et al 1994), but maternal height was also a predictor of birth weight. This is not unexpected since short stature may well be a re¯ection of poor maternal nutrition
during her growing years. Our data are consistent with these observations. Maternal haemoglobin concentration appeared to have two distinct in¯uences on birth outcome. At booking, haemoglobin concentration was positively associated with newborn size, namely weight, length and head circumference. This association in early pregnancy might re¯ect overall maternal nutritional status (Godfrey et al, 1991). However, the fall-off in ponderal index at haemoglobin concentrations exceeding 12.5 g/dl might point to maternal circulatory dysfunction. In the second and third trimesters, haemoglobin was negatively associated with size at birth. We speculate that this re¯ects the well recognized relationship between maternal hemodilution and fetal growth. Hemodilution is a feature of normal pregnancy, and haemoglobin concentrations exceeding 12.5 g/dl indicate inadequate hemodilution. Rosso et al (1992) suggest that women who are chronically undernourished are unable to haemodilute adequately, and their reduced plasma volume is associated with a lower mean birth weight (Goodlin, 1983). Failure to haemodilute appropriately, as indicated by higher minimum haemoglobin during pregnancy, is associated with smaller size at birth (Goodlin, 1983). Lastly, anaemia is likely to obscure these relationships. Anaemia in pregnancy is, itself associated with smaller size at birth (Kramer, 1987). In our population of women, haemoglobin concentration below 10.5 g/dl in the third
Maternal nutritional status M Thame et al
138
trimester was associated with lower birth weight, ponderal index, and a higher placental to birth weight ratio. In our sample, age was also related to foetal outcome. Younger mothers have smaller babies for a variety of reasons. One contributor is thought to be competition for nutrients in the younger mothers who are themselves still growing. Younger mothers in our population were also thinner, and had lower parity, and the effect of age was not related to parity. However, age did not make an independent contribution to birth weight after taking account mother's weight, height, parity, child's gender and gestational age. Similar ®ndings have been seen in previous studies where there were no signi®cant differences in birth weight between adolescent and young adult women (Horon et al, 1983). In summary lighter, shorter babies and babies who were disproportionate were born to mothers who were younger, lighter, shorter and thinner. Inadequate hemodilution, and anaemia probably also contributed to smaller size at birth, but these inferences were based on haemoglobin concentrations only. Because variations in size at birth which re¯ect fetal growth restraint appear to be early markers of chronic cardiovascular disease in adulthood, there is the need to explore the mechanisms which underlie these apparent effects of maternal nutritional status on fetal growth and development. References Baqui AH, Arifeen SE, Amin S & Black RE (1994): Levels and correlates of maternal nutritional status in urban Bangladesh. Eur. J. Clin. Nutr. 48, 349±357. Barker DJP (1994). Mothers, Babies and Disease in Late Life. BMJ Publishing Group: London. Barker DJP, Gluckman PD, Godfrey KN, Harding JE, Owens JA & Robinson JS (1993): Fetal nutrition and cardiovascular disease in adult life. Lancet. 341, 938±941. Barker DJP, Meade TW, Fall CHD Lee A, Osmond C, Phipps K, et al. (1992): Relation of fetal and infant growth to plasma ®brinogen and factor VII concentrations in adult life. Br. Med. J. 304, 148±152. Chow BF, Lee C-J (1964): Effect of dietary restriction of pregnant rats on body weight gain of the offspring. J. Nutr. 82, 10±18. Cooper R, Rotimi C, Ataman S, McGee D, Osotimehin B, Kadiri S, et al (1996): Hypertension prevalence in seven populations of African origins. Amer. J. Public Health. (in press). Forrester TE, Wilks RJ, Bennett FI, Simeon D, Osmond C, Allen M, et al (1996): Fetal growth and cardiovascular risk factors in Jamaican schoolchildren. Br. Med. J. 312, 156±160. Godfrey KM, Redman CWC, Barker DJP & Osmond C (1991): The effect of maternal anaemia and iron de®ciency on the ratio of fetal weight to placental weight. Br. J. Obstet. Gynecol. 98, 886±891.
Godfrey KM, Forrester T, Barker DJP, Jackson AA, Landman JP, Hall J.St E, et al (1994): Maternal nutritional status in pregnancy and blood pressure in childhood. Br. J. Obstet. and Gynaecol. 101, 398± 403. Goodlin RC, Dobry CA, Anderson JC, Woods RE & Quaife M. (1983): Clinical signs of normal plasma volume expansion during pregnancy. Am. J. Obstet. Gynecol. 145, 1001±1009. Gueri M, Jutrsm P & Sorhaindo B (1982): In Pregnant Women: a reference table of weight-for-height by week of pregnancy. Am. J. Clin. Nutr. 35, 609±611. Horon I, Strobino D & MacDonald H (1983): Birth weights among infants born to adolescent and young adult women. Am. J. Obstet. Gynecol. 146, 444±449. Kramer MS (1987): Determinants of low birth weight: Methodological assessment and meta-analysis. Bulletin. Bull. WHO. 65, 663±737. Langley SC & Jackson AA (1994): Increased systolic blood pressure in adult rats induced by foetal exposure to maternal low protein diet. Clin. Sci. 86, 217±222. Law CM, Barker DJP, Bull AR & Osmond C (1991): Maternal and fetal in¯uence on blood pressure. Arch. Dis. Child. 66, 1291±1295. Levy L & Jackson AA (1993): The modest restriction of dietary protein during pregnancy in rat: fetal and placental growth. J. Dev. Physiol. 19, 113±118. Mavalankar DV, Gray RH, Triedi CR, Parkh VC (1994): Risk factors for small for gestational age births in Ahmedadab, India. J. Trop. Pediatr. 40, 285. McCance RA, Widdowson EM (1974): The determinants of growth and form. Proc. R. Soc. Lond. [Biol]. 185, 1±17. Miall WE, Del Campo E, Fodor J, Nava Rhode JR, Ruiz L, Standard KL, et al (1972): Longitudinal study of heart disease in Jamaican rural population: prevalence, with special reference to ecg ®ndings. Bull. WHO. 46, 429±442. Naeye RL (1979): Weight gain and the outcome of pregnancy. Am. J. Obstet. Gynecol. 135, 3±9. Pivalizza PJ, Woods DL, Sinclair-Smith CC, Kaschula ROC & Pivalizza EG. (1990): Placentae of Light for Dates Infants Born to Underweight Mothers at Term: A Morphometric Study. Placenta. 11, 135± 142. Poppitt SD, Prentice AM, Goldberg GR, Roger S & Whitehead RG. (1994): Energy-sparing strategies to protect human fetal growth. Am. J. Obstet. Gynecol. 171, 118±125. Rosso P, Donoso E, Braun S, Espinoza R & Salas S. (1992): Hemodynamic Changes in Underweight Pregnant Women. Obstet. Gynaecol. 79, 908±912. Stein ZA & Susser MW (1975): The Dutch famine 1944±45 and the reproductive process I. Effects on six indices at birth. Paediatr. Res. 9, 70±76. Thongprasert K, Tanphaichitre V, Valyasevi A, Kittigool J & Durnin JVGA (1987): Energy requirements of pregnancy in rural Thialand. Lancet. ii, 1010±1012. Widdowson EM & McCance RA (1963): The effect of ®nite periods of undernutrition at different ages on the composition and subsequent development of the rest. Proc. R. Soc. Lond. [Biol]. 158, 329±342. Widdowson EM (1971): Intra-uterine growth retardation in the pig. 1. Organ size and cellular development at birth and after growth to maturity. Biol. Neonate. 19, 329±340. Zeeman FJ & Stanbrough EC (1969): Effect on maternal protein de®ciency on cellular development on the fetal rat. J. Nutr. 99, 274±282.
