Bull. Org. mond. Sante 1962, 26, 203-211Bull. Wld Hlth Org.

Effect of Nutrition on Pregnancy and Lactation *C. GOPALAN, M.D., Ph.D.'

Pregnancy and lactation constitute states of considerable physiological stress whichimpose increased nutritional demands. If these demands are not adequately met, it may beexpected that not only the nutritional status of the subject will be affected, but also thecourse ofpregnancy and lactation. While a great deal of work with experimental animalshas been carried out to elucidate the role of nutrition in pregnancy and lactation, thequestion arises how far these experimental results are applicable to human subjects. Theunfortunate nutritional situation prevalent in certain under-developed countries affordsopportunities for the study ofthe effects ofnutritional deficiencies on the course ofpregnancyand lactation in the human subject. In this paper, the available literature on the effect ofmaternal nutrition on the course ofpregnancy and the condition of the infant at birth isreviewed, as is the effect of the state of maternal nutrition on the output and chemicalcomposition of milk in nursing mothers. The review reveals many important gaps in ourknowledge and highlights the need for further work on this important problem.

Expectant and nursing mothers and infants andchildren constitute the most vulnerable segments ofa population from the nutritional standpoint. Inany consideration of the problems of undernutrition,these segments require special consideration. Theunfortunate nutritional situation prevalent in manyunder-developed countries of the world today offersopportunities for studies on the effect of maternalmalnutrition on the course of pregnancy, the condi-tion of the infant at birth, and the state of lactationof the mother. Such studies may also be expectedto throw light on maternal nutritional requirementsduring pregnancy and lactation.

NUTRTION AND PREGNANCY

The important questions which require to beconsidered under this head are:

1. How does the state of maternal nutrition affectthe course of pregnancy?

2. How does the state of maternal nutrition affectthe condition of the infant at birth and the neonatalperiod?

* Paper submitted to the Joint FAO/WHO Expert Com-mittee on Nutrition, April 1961.

1 Director, Nutrition Research Laboratories, Hyderabad,India.

3. What are the optimal maternal nutritionalrequirements in pregnancy which would ensureproper health of the mother, normal course of herpregnancy and satisfactory condition of her infantat birth?

The course of pregnancy

Pregnancy wastage. A great deal of informationis available pointing to the effect of nutrition on thecourse of pregnancy in experimental animals (e.g.,Nelson & Evans, 1946, 1953). Nelson & Evans(1953) observed that in rats, when the dietary proteinlevel was reduced to less than 5 %, the incidence offoetal resorptions was as high as 70%-100%. Thecritical need for protein was apparent only duringthe earliest stages of placental and foetal develop-ment. The precise practical significance of theseobservations from the point of view of the humansubject, however, requires cautious appraisal. Asurvey carried out in South India revealed thatamong poor women whose dietaries during preg-nancy provided 1400-1500 calories and about 40 gof protein daily, nearly 20%. of pregnancies hadterminated in abortions, miscarriages or stillbirths.2This figure may well be an underestimate and maynot include abortions in the early stages of preg-

2 Gopalan, C. Address to the 1960 Annual Meeting ofthe American Dietetic Association (to be published).

1095 - 203 -

204 C. GOPALAN

nancy. In a recent study among poor women inHyderabad, Kalpakam Shankar1 found that theincidence of pregnancy wastage was about 16%.Such high incidence of pregnancy wastage appearsto be the usual finding among undernourished com-munities in other parts of India also. To what extentthis is attributable to lack of proper obstetrical care,is, however, a moot question. The possible role ofmalnutrition in contributing to the high incidence ofpregnancy wastage in undernourished communitieswould seem to need consideration.

Gain in body-weight during pregnancy. The rela-tively low gain in body-weight of undernourishedwomen during pregnancy has been commented uponby some workers (Clements & Bocobo ;2 Venkata-chalam et al., 1960; Gopalan 3). While the usualweight gain during pregnancy among well-fed com-munities has been reported to be around 25% of theinitial weight, Indian women of the poor socio-economic group weighing initially around 42 kgwere found to gain on an average only about 6 kgduring pregnancy.The significance of the comparatively small gain

in body-weight of undernourished women duringpregnancy requires consideration. In -pregnancythere are obviously profound changes in body com-position. In a study of the body composition ofeighteen undernourished pregnant women in SouthIndia (Venkatachalam et al., 1960), it was observedthat between the 9th and the 14th weeks of gestationthe total body water was about 56% of the body-weight, a value which corresponded to that observedin the non-pregnant women of the same population.Between the 20th and the 28th weeks of gestation,the total body water was about 66% and beyond the28th week about 70%. The interesting finding wasthat the increase in body-weight of the subjects whichwould have been expected from the absolute in-crement of total body water was much greater thanthe actual increase in body-weight. Since nitrogenbalance studies in these subjects indicated that theywere in positive nitrogen balance on their usualprotein intake, it was concluded that these womenwere actually losing body fat during pregnancy. Thelower weight gain of these undernourished mothersas compared to well-fed mothers may be a reflectionof smaller gain of body protein. Similar studies of

1 Data to be published.'Unpublished report to WHO on nutrition in Ceylon,

1957.Address to the 1960 Annual Meeting of the American

Dietetic Association (to be published).

body composition of well-fed pregnant women mayserve to decide this question, and may also serve toexplain some of the contradictory findings withregard to the relationship between maternal body-weight changes during pregnancy and the birthweight of the infant.The significance of the loss of body fat during

pregnancy observed in the study reported aboverequires elucidation. The question is whether suchloss of body fat is the result of undernutrition. Inthis connexion, the observations of Scitchik & Alper(1956) are interesting. These workers carried outstudies of body composition in well-fed Americanpregnant women and concluded that the increase inlean body tissue in pregnancy was associated with adepletion of body fat, the fat being probably utilizedfor the synthesis of lean body tissue.

There is obviously a need for further studies ofbody composition in pregnancy in different socio-economic groups.

Toxaemias of pregnancy. A high incidence ofeclampsia has been observed among pregnant womenof the low socio-economic group in certain parts ofIndia. It has been claimed from time to time thatmalnutrition may play a role in the development oftoxaemia of pregnancy. The evidence for such aclaim, however, needs careful evaluation.Summing up an extensive survey of available data

regarding the incidence of toxaemia in different partsof the world, Dieckmann (1952) observed: "Eclamp-sia, toxaemia, hyperemesis and abruptio placentaeare either unknown or are very rare among the nativewomen of Kenya, Uganda, Zululand, Tanganyika,Belgian Congo, Ethiopia, Persia, Java, Hawaii, Brit-ish Malaya, Alaska, 'Australia and the date oases ofAfrica, where habits and diet have not been changedby the White race. In contrast, eclampsia is verycommon in Algiers, Cape Town, Colombo andPuerto Rico, whose natives have adopted many ifnot all of the diet and other habits of the Whiterace." This may well be an over-simplification ofthe problem.The reported incidence of pre-eclampsia and

eclampsia in different parts of the world has beenset out in Table 1. It will be noted that through theyears there has been a reduction in the incidence ofeclampsia in Europe, the USA and India, while theincidence of pre-eclampsia has not shown significantvariation. The studies of Ebbs et al. (1941) and of thePeople's League of Health (1942) failed to show anysignificant effect of diet on the incidence of toxaemia.Nelson (1955) studied the incidence of eclampsia

EFFECT OF NUTRITION ON PREGNANCY AND LACTATION 205

TABLE IINCIDENCE OF TOXAEMIAS OF PREGNANCY

IN DIFFERENT COUNTRIESa

Pre- EclampsiaCountry Year eclampsia ( %

I Incidence)incidence)

British Isles 1938-40 9.3 1.041947-50 11.5 0.632

Australia 1939-41 8.90 0.7091947-50 11.00 0.390

New Zealand 1936-40 6.3 0.51947-49 7.0 0.56

USA:Boston Lying-in Hospital 1948-49 5.4 0.81Margaret Hague Hospital,New York 1947-49 8.1 0.09

Chicago Lying-in Hospital 1947-50 8.5 0.84Jackson MemorialHospital 1956-57 7.43 0.113

India:Women and Children's 1936-38 4.4 2.1

Hospital, Madras 1955-57 7.4 1.71959-60 7.8 1.2

a Data supplied by M. K. Krishna Menon (personal communi-cation).

and pre-eclampsia among different socio-economicgroups in Aberdeen. Diet surveys by Thomson(1959) show that the diets of these various socio-economic groups differed in quantity and quality.Despite such differences in the dietaries, Nelson'sstudies revealed that the incidence of pre-eclampsiawas the same in all social classes. Nelson concludedthat (a) the incidence of pre-eclampsia had notaltered through the years; (b) the incidence ofeclampsia had diminished considerably, and (c) theincidence of pre-eclampsia was the same in all classes.Dawson (1953) from his studies in Australia observedthat while good antenatal care had almost eliminatedeclampsia, it had not reduced the incidence of pre-eclamptic toxaemia.A careful appraisal of the available evidence thus

fails to provide convincing proof that malnutritionis a significant factor contributing to pre-eclampsia.However, the high incidence of eclampsia in under-nourished populations raises two possibilities:(a) that malnutrition may in some manner aggravatethe course of pre-eclampsia, or (b) that the higherincidence of eclampsia in undernourished commu-

nities is largely the result of poor obstetric care, asa result of which cases of pre-eclampsia are notspotted early and treated adequately with the con-sequence that a high proportion of these develop intoeclampsia. Obstetricians seem to be almost evenlydivided between these two schools of thought. Incommunities with primitive obstetric care and poornutrition, it should be possible to devise field exper-iments which would serve to clarify the position.

The condition of the infantBirth weight and incidence ofimmaturity. There is

now ample evidence pointing to the effect of maternalnutritional status on the birth weight of the infant.In a survey carried out in South India, the meanbirth weight of infants of the low socio-economicgroup was found to be about 2.8 kg while that ofinfants of the high socio-economic group was about3.1 kg.' This difference was found to be statisticallysignificant. If all infants including those who werenot born at full term had been included, the differ-ence between the two groups would have been evenwider.

Using a birth weight of 2.5 kg as the criterion, itwas found that the incidence of immaturity among theinfants of the low socio-economic group was nearly30% while the corresponding figure among the highsocio-economic group was about 14 %. If a birthweight of 2.0 kg was used as the criterion, the in-cidence of immaturity was less than 2% amonginfants born to mothers in the high socio-economicgroup as compared to 10 % in the poor socio-economic group.What is the practical significance of the relatively

low birth weight of infants of undemourished moth-ers? It may be safely conceded that a birth weightof less than 2.5 kg, or even 2.0 kg, need not neces-sarily indicate functional inadequacy. Indeed, mostinfants with such low birth weights were observedto thrive normally. It would appear, however, thatwhile birth weight may not be a reliable clinicalcriterion in individual cases, it may be a useful yard-stick when whole communities or population seg-ments are taken into consideration. For example, inMadras City in South India, the infant mortalityrate was 136 per 1000 live-births in the year 1954.Nearly 25% of these deaths occurred within sevendays after birth and 40% within the first month.This neonatal mortality was found to contributethe bulk of infant mortality in this area. An analysis

1 Gopalan, C. (1949) Unpublished data.

206 C. GOPALAN

of the causes of neonatal deaths revealed that nearly73% of all these deaths occurred in infants withbirth weights of less than 2.0 kg. It would thusappear that birth weight in a community is an indexof neonatal viability and that maternal nutritionalfactors which result in low birth weight also contrib-ute to the high infant mortality observed in under-developed regions.

Congenital malfornations. It is now well estab-lished that congenital malformations can be inducedin the offspring of mammals by maternal dietarydeficiencies. But congenital anomalies due to sponta-neously occurring maternal nutritional deficiencyhave not been recognized in the human being. Theavailable literature lends no support to the view thatthe incidence of congenital malformations in infantsof malnourished mothers is significantly differentfrom that in infants of well-fed mothers. The strictconditions and devices necessary to produce con-genital anomalies in experimental animals have prob-ably no counterpart in human reproduction. Incontrast to the rat, the organogenetic period inwhich most malformations are determined is rel-atively short in the human subject; and it is followedby a long period of foetal growth in which damagedfoetuses can be eliminated. The incidence of preg-nancy wastage among malnourished mothers mayindeed be a reflection of such elimination of damagedfoetuses.The syndrome of endemic cretinism may be justifi-

ably looked upon as a manifestation of the effect onthe offspring of maternal dietary deficiency of iodine,though it is possible that other factors may alsocontribute to its etiology. The results achieved inSwitzerland in the matter of eradication of endemiccretinism through iodization of salt and otherdietary improvement may be considered a con-vincing demonstration of the possible effect ofmaternal diet on the physical and mental make-upof the infant.Warkany et al. (1959) have reported congenital

anomalies following the ingestion of aminopterin bythe mother during early pregnancy. This observa-tion brings to light the possible danger of congenitalanomalies resulting from the use of potent anti-metabolites during pregnancy.

Effect of maternal malnutrition on foetal storage ofnutrients. The three major nutritional deficiencydisorders in infancy and childhood encountered inmany under-developed countries are protein mal-nutrition, hypovitaminosis A and anaemia. It isimportant to consider to what extent maternal mal-

nutrition during pregnancy may contribute to thedevelopment of these deficiencies in infancy andchildhood.

It has been found that among poor pregnantwomen in South India, the vitamin A concentrationin the serum declined from 104 IU in the first tri-mester to 68 IU in the third trimester.' After deliv-ery, there was again a rise of serum vitamin A con-centration. The serum carotene, however, actuallyshowed a slight increase with advancing pregnancy.These observations are somewhat in line with thoseof other workers (Lund & Kimble, 1943). Thevitamin A of cord blood from infants of thesemothers was about 50 IU per 100 ml. On the otherhand, cord blood obtained from a group of infantswhose mothers had received 10 000 IU of vitamin Aorally for a few days in the last trimester returnedvalues averaging 85 IU per 100 ml.The above observations would indicate that cor-

rection of dietary deficiency of vitamin A in themother during pregnancy might significantly improvethe nutritional status of the infant with regard tovitamin A. The available literature with regard toplacental transfer of vitamin A and carotene revealsconsiderable differences in the behaviour of differentspecies. In the rat the amount of vitamin A whichpasses into the foetal livers appears small and in-dependent of the vitamin A fed to the mothers(Dann, 1932, 1934). In calves, on the other hand,the vitamin A content of the maternal diet has beenfound to determine the amount of vitamin A trans-ferred to the foetal liver. Thus, while calves ofundosed cows had 2.5 IU of vitamin A per g of liver,those of vitamin-A-supplemented cows had valuesas high as 142 IU (Walker et al., 1949). In pups,kittens, lambs, goats and pigs also it has been dem-onstrated that the foetal storage of vitamin A couldbe greatly improved by dosing the pregnant motherwith vitamin A (Thomas et al., 1947).Data with regard to the vitamin A content of livers

of newborn infants among malnourished commu-nities are unfortunately rare. Woo & Chu (1940)have reported that full-term infants in China showedvalues for hepatic vitamin A as low as 10 IU/g,while those of the USA (Lewis et al., 1943) andFinland (Skurnik et al., 1944) had a liver vitamin Acontent as high as 134 IU/g and 154 IU/g respec-tively. It is reasonable to expect that maternal mal-nutrition with respect to vitamin A, because of theresultant poor foetal hepatic storage of vitamin A,

1 Venkatachalam, P. S., Bhavani Belavady & Gopalan, C.Data to be published.

EFFECT OF NUTRITION ON PREGNANCY AND LACTATION 207

may predispose the infant to vitamin A deficiency inlate infancy and early childhood.Another factor which may also be expected greatly

to influence the hepatic storage of vitamin A in theinfant is the pattern of feeding in the immediateneonatal period. It has been found that the vitaminA content of colostrum even among malnourishedmothers is as high as 600 IU/100 ml. The vitamin Aconcentration in milk gradually declines to about70 IU/100 ml. If the infant is put to breast imme-diately after birth, as is the case among many under-fed communities, better hepatic storage may be ex-pected than if the practice of not feeding the infantat the breast for the first three days-a habit in voguein some other communities-is followed. Calves notallowed access to colostrum have been found to diewithin a few days. In such animals, the hepaticvitamin A storage has often been found to be toolow to be measurable (Moore, 1957).

Field studies to determine how far the pattern ofinfant feeding in the early neonatal period influencesthe onset and course of vitamin A deficiency inchildren among different poor communities may yieldvaluable information.The possible effect of anaemia in pregnancy on the

haematological status of the infant at birth and inthe neonatal period also deserves consideration. Ina survey of the haemoglobin levels of nearly 400pregnant women in South India, Kalpakam Shankarlfound a progressive increase in the incidence andseverity of anaemia with the advance of pregnancy.Thus, while 20% of subjects had haemoglobin levelsless than 10 g% in the first trimester, over 50% hadsuch low values in the later stages of pregnancy.Among the infants, the haemoglobin values de-creased from about 18-20 g% at birth to about10 g% by about the third month. After the sixthmonth, there was a further deterioration and by the18th month the average haemoglobin level wasaround 8.5 g%. The studies of Bhavani Belavady& Gopalan (1959) had revealed that the iron contentof milk of these mothers in the earlier stages oflactation was not lower than the values reportedelsewhere, and further, that there was no correlationbetween haemoglobin levels of the nursing mothersand the iron content of their milk samples. In viewof this, the high incidence of anaemia within thefirst six months may be attributable to inadequatefoetal stores of iron, probably reflecting the unsatis-factory state of maternal nutrition with regard to

1 Data to be published.

iron. It has been stated that foetal iron stores aredeposited in the last trimester of pregnarfcy and thatthey are influenced by the mother's nutrition withrespect to iron during pregnancy (Darby, 1950).The normal full-term infant born of a non-anaemicmother is believed to possess adequate iron storesat birth, so that a dietary supplement is not requiredfor several months. However, conclusions regardingfoetal storage of iron and its adequacy for the forma-tion of new haemoglobin in the full-term infant arenot in complete agreement. Early studies indicatedthat these iron stores compensate for the meagre ironcontent of milk. Later studies, however, show thatthe stores of non-haemoglobin iron in the newborninfant's liver are not as large as had been assumed(Stearns & McKinley, 1937; Smith et al., 1950).Gladstone (1932), on the basis of chemical andmicroscopic studies, saw no evidence of large irondeposits in the liver in the last four months of gesta-tion, the maximum iron storage being reached from1 to 10 weeks after birth. Langley (1951) confirmedthis observation and showed that there was a progres-sive siderosis in the liver and spleen, starting threedays after birth and reaching a maximum during thethird week of life. The recession in extramedullaryhaemopoiesis and red blood cell haemolysis in theneonatal period are believed to be factors whichincrease neonatal iron stores. If, therefore, the neo-natal iron store is derived to a great extent fromhaemoglobin iron, the question arises whether thecontent of haemoglobin iron in the newborn infantsof undernourished mothers is inadequate. Studieson newborn infants of the poor socio-economicgroup (Jayalakshmi et al., 1957) had not revealedabnormally low haemoglobin concentration in theseinfants. But in the absence of data regarding theblood volume of these infants, no precise conclusionabout the total haemoglobin iron is possible. Thespeed with which these infants develop anaemia inthe neonatal period, is, however, highly suggestiveof defective iron storage in the foetal or neonatalperiod.

Maternal nutritional requirements in pregnancyVarious national bodies have made recommenda-

tions regarding nutrient allowances in pregnancy.The scientific basis underlying some of these recom-mendations would seem to need careful evaluation.The same observations would also apply to therecommendations regarding nutrient allowances inlactation. The observations discussed above doindicate that the inadequate dietaries of pregnant

208 C. GOPALAN

women in some under-developed countries arereflected in the condition of the mother as well as ofthe infant. While this is understandable, what isintriguing is the fact that a large number of thesemothers are able to go through pregnancy apparentlynormally and to deliver normal infants of averagebirth weights. Furthermore, most of these mothersare able to breast-feed their infants successfully overlong periods. This raises some interesting and impor-tant questions connected with the adaptive mecha-nisms brought into play during pregnancy andlactation to meet the increased nutritional demandsof the mother. A great deal of further work isnecessary to understand these mechanisms; andunless this is done a rational approach to the problemof maternal nutritional requirements during preg-nancy and lactation would not be possible.

NUTRITION AND LACTATION

A descriptive account of the available data on thissubject has been presented elsewhere (Gopalan &Bhavani Belavady, 1960). Some of these data maybe highlighted here and some further observationsmay be briefly discussed.

Studies of the output of breast milk at differentstages of lactation reveal that, in spite of their in-adequate diets, women of the low socio-economicgroup put out 400-600 g of milk daily for periodsextending to over a year. Studies of the chemicalcomposition of milk show that as far as proximateprinciples are concerned, the milk from these mothersis nearly as satisfactory as those from well-fedmothers. The concentration of the vitamins,however, appears to be rather low in the milksamples from undernourished mothers. Apparently,the inadequate dietaries of these mothers are reflectedin the concentrations of vitamins in their milk ratherthan in that of proximate principles. This conclusionis, however, applicable only if the diets supply atleast between 45 g and 60 g of protein daily. It ispossible that lower levels of protein intake may bereflected in the low concentration of protein inthe milk.

Effect ofprotein supplementationDietary protein supplementation to the mother has

not been shown to bring about any significantincrease in protein concentration of the milk, wherethe diets already contained 50 g of protein daily.However, protein supplementation has been foundto have an effect on two biochemical features of the

milk; the significance of these effects remains to beelucidated.The creatine and creatinine concentration of the

milk of poor Indian mothers was found to be veryhigh as compared with figures reported from theUSA. Thus, Bhavani Belavady (1959) observed thatthe creatinine N2 was 2.3 mg/100 ml and creatineN2 3.4 mg/100 ml in milk samples from poor SouthIndian mothers. Karmarkar et al. (1959) in WesternIndia observed values as high as 2.9 mg/100 ml and8.4 mg/100 ml for creatinine N2 and creatine N2respectively. These values are far higher than thosereported by Macy & co-workers (1949), whoobserved a creatine N2 concentration of 1.0 mg/100ml and creatinine N2 concentration of 1.0 mg/100 mlin milk samples from American women. Gopalan &Bhavani Belavady (1960) found that dietary proteinsupplementation to the mother brought about asignificant reduction in the abnormally high creatinevalues in milk; the creatinine concentration, however,was apparently unaffected. Creatine in smallamounts was excreted in the urine of the subjects.Protein supplementation had, however, no effect oncreatine and creatinine output in urine. The serumcreatine and creatinine figures in these subjects werenormal. The significance of the high creatine con-tent in milk and the effect of protein supplementationin bringing about a reduction thereof require furtherelucidation.Another interesting effect of protein supplemen-

tation in these mothers was on the xanthine oxidaseactivity. Bhavani Belavady 1 observed that thexanthine oxidase activity of milk in these motherswas increased significantly after protein supplemen-tation from 12.4 ,u 02 consumed to 28.5/,l °2 con-sumed per ml of milk per hour.

Thus, though protein supplementation had nosignificant effect on protein concentration in milk,it may be noted that it exerted a significant effecton two biochemical constituents of milk. Further,it was found that protein supplementation broughtabout a significant increase in body-weight and inserum albumin concentration of these mothers.These factors may have to be considered in assessingthe adequacy of the dietary protein level in thesecases.

Casein: whey protein ratio in milk

Lutz & Platt (1958) had reported that the casein:whey protein ratio in milk obtained from mal-

1 Data to be published.

EFFECT OF NUTRITION ON PREGNANCY AND LACTATION 209

nourished Indian mothers was high as comparedwith the ratio in English mothers. Thus, theseauthors observed a ratio of 1.30 in Indian mothersand 1.08 in English mothers. On the other hand,studies carried out by Bhavani Belavady & Gopalan(to be published) showed that among poor Indianwomen in Coonoor, low values of about 0.3 wereobtained for the casein: whey ratio. In Hyderabad,however, using the same techniques, values of about1.08 were observed. The reasons for this strikingdifference are not clear. Among the women ofHyderabad in whom the effect of protein supple-mentation was studied, the casein:whey ratio wasfound to be unaltered by protein supplementation.It must, however, be pointed out that these motherswere not so malnourished as those evidently investig-ated by Lutz & Platt. There is obviously furtherscope for work on this interesting aspect.

Vitamins

Bhavani Belavady & Gopalan (1960; and datato be published) found that dietary supplementationwith riboflavin, thiamine and ascorbic acid broughtabout a significant increase in the concentration ofthese vitamins in milk. On the basis of these sup-plementation studies, the authors concluded thatthe minimum dietary intake of riboflavin andascorbic acid necessary to maintain a maximalconcentration of these vitamins in milk was of theorder of 3.0 mg and 200 mg respectively.With regard to vitamin A, however, it was

observed that dietary supplementation with vitaminA was not reflected in the concentration of thevitamins in milk even when the supplementationwas maintained for 6-9 months. However, it wasfound that in mothers who became pregnant in thecourse of lactation, dietary supplementation withvitamin A had a profound effect on vitamin A con-centration in milk. Thus, in two such cases it wasnoticed that the vitamin A concentration in milk rosefrom 60 IU/100 ml to 530 IU/100 ml and from75 IU/100 ml to 240 IU/100 ml after four weeks ofsupplementation with 5000 IU of vitamin A daily.If this interesting observation is confirmed by moreextended work, it would mean that the metabolicchanges in pregnancy facilitate transfer of vitamin Aacross the mammary gland. To elucidate the possiblerole of oestrogens in facilitating the transfer ofvitamin A to milk, some studies are currently inprogress in the Nutrition Research Laboratories,Hyderabad.

TABLE 2

RESUMPTION OF MENSTRUATION AFTER DELIVERY INSOUTH INDIAN MOTHERS

Percentage of mothers who hadAge of child resumed menstruation(months)

Poor Well-to-do

1-3 0 22

4-6 8 84

6-9 25 88

10-12 45 100

13-15 47 100

16-18 70 100

MineralsThe concentration of calcium and iron in the milk

samples obtained from women of the poor socio-economic group in South India were found to bearound the reported normal range in other partsof the world (Bhavani Belavady & Gopalan, 1959).This is interesting considering the fact that the dietsof these mothers supplied hardly 300 mg of calciumdaily and the majority of the mothers were anaemic.Supplementation of the diet with calcium and irondid not bring about an increase in the concentrationof these nutrients in milk; indeed, paradoxicallyenough, there was actually a fall in the concentrationof these nutrients. This point has been discussed inearlier communications (Bhavani Belavady & Go-palan, 1959, 1960).

Resumption of menstruation during lactationA recent study 1 has revealed interesting differences

between undernourished and well-fed mothers withregard to the time-interval between delivery andresumption of menstruation. The data are set outin Table 2. It will be seen that menstruation wasresumed much later in the majority of the poormothers than in the well-to-do. The significance ofthis finding, is, however, hard to assess. Whetherearly resumption of menstruation in the well-fedmothers is the result or the cause of their failure ofsuccessful lactation remains to be elucidated.

DiscussionThe observations discussed above show that the

nutritional status of the mothers is reflected in the

1 Bhavany Belavady. Data to be published.

C. GOPALAN

TABLE 3INTAKE OF MILK BY SOUTH INDIAN BABIES a

Milk intake(ml)IinF24 hours by babiesDay after delivery with birthLweight of:

4-5 pounds 8-9 pounds

2nd 80 80

3rd 200 310

4th 260 400

5th 310 430

6th 330 450

7th 360 510

8th 350 540

9th 370 540

10th 360 540

a Reproduced, by permission, from Athavale (1960).

chemical composition of the milk, at least withregard to the concentration of vitamins. It will beimportant to know to what extent the growth ofinfants in these communities could be improvedthrough supplementation of the maternal diet withvitamins. The growth-curve of the infants in these

R1S

Les femmes enceintes et les meres allaitantes, les nour-rissons et les jeunes enfants constituent les secteurs lesplus vulnerables de la population, du point de vue ali-mentaire. En effet, des dommages irreparables peuventsurvenir si les exigences alimentaires accrues durant cesperiodes de la vie ne sont pas satisfaites. La sous-alimentation, dans plusieurs pays insuffisamment deve-loppes, permet d'etudier 1'effet de la malnutrition mater-nelle sur le cours de la grossesse, l'etat de l'enfant i lanaissance, la lactation, et de pr&iser les exigences ali-mentaires en rapport avec la physiologie de la femme.En s'appuyant sur des observations anterieures, sur

les experiences faites chez l'animal, et sur ses propresrecherches, dans l'Inde, l'auteur aborde les questionssuivantes: l'influence de l'alimentation sur le cours dela grossesse et l'etat du nouveau-ne; les exigences ali-mentaires qui assureraient une grossesse menee a termeet la naissance d'un enfant ayant de bonnes chances deviabilite. La proportion d'avortements naturels dans lespopulations sous-alimentees de l'Inde est estimee A 16-20 %. La part due au manque de soins obstetriques, dans

poor communities is nearly parallel to that of infantsfrom the better-fed groups, at least in the first fewweeks, but the initial difference in body-weight ismaintained. Perhaps supplementation of the ma-ternal diet with vitamins may actually help to over-come this difference. It may be possible to investi-gate this question through carefully controlledexperiments. The work of Widdowson (quoted byGyorgy, 1960) has indicated that the initial growthretardation resulting from undernutrition in the neo-natal period may be irreversible. If this is so, it willbe important to explore the ways by which the growthrate of these infants in the early neonatal periodcould be improved.

In this connexion, the observations of Athavale(1960) appear interesting; they suggest that theamount of breast milk consumed by the infant in24 hours may be determined by the body-weightof the infant, as shown in Table 3. This requires tobe confirmed by extended studies as its implicationsappear important.The foregoing account shows that there are still

important areas in which further work is necessaryto elucidate the precise role of nutrition in pregnancyand lactation. Such work deserves high priority inview of the widespread maternal and infant malnu-trition prevalent in many under-developed regionsof the world today.

ce chiffre, n'est pas connue; celle qui revient a la mal-nutrition devrait etre evalu&e. Le gain en poids de lamere pendant la grossesse, qui est de 1/4 du poids initialdans les collectivites bien nourries, est d'environ 1/7dans les collectivites mal nourries; cette diff6rence sembledue a une plus grande deperdition de proteines, d'eauet de graisse du corps dans le second groupe.Quant A la preeclampsie, elle ne parait pas dependre

directement de la malnutrition. Toutefois, pour expliquerla frequence elev6e de l'&clampsie dans les collectivitesmal nourries, on peut avancer que a) la malnutritionaggrave l'evolution de la preeclampsie, ou que b) fautede soins obstetricaux, la pre6clampsie n'est pas deceleeou traitee a temps, et qu'elle 6volue le plus souvent eneclampsie. Les obstetriciens sont eux-memes divises surce point.

Il existe une diff6rence significative entre le poids A lanaissance des enfants mis au monde dans les collectivitesaisees et dans celles qui sont d6favoris&es. La differenceserait encore plus grande si l'on tenait compte des avorte-ments naturels. L'insuffisance de poids A la naissance

210

EFFECT OF NUTRITION ON PREGNANCY AND LACTATION 211

(moins de 2,5 kg) est de l'ordre de 30% des naissancesdans les groupes socio-economiquement faibles et de14% dans les groupes plus favorises. Cette proportionest respectivement de 10% et 2% si l'on applique lecritere de 2 kg de poids initial. Si l'on peut admettre,dans les cas individuels, qu'un poids a la naissance de2-2,5 kg n'indique pas necessairement F'inviabifite, cechiffre donne une indication utile lorsque l'on considerela mortalite neonatale dans l'ensemble d'une population.A Madras, par exemple, la mortalite neonatale intervientpour une large part dans la mortalite infantile, qui etaitde 1360/oo en 1954; I'analyse des causes de mortaliteneonatale montre que 73 % des deces survenaient chezdes enfants qui pesaient moins de 2 kg a la naissance.Dans une collectivite, le poids a la naissance est unindice de viabilite neonatale, et les facteurs nutritionnelsqui concourent a un poids insuffisant du nouveau-neconcourent aussi a elever la mortalite infantile dans lesregions sous-developpees.Des malformations congenitales dues A la malnutrition

maternelle n'ont pas et signalees dans l'espece humaine.La frequence des malformations congenitales ne paraitpas plus elevee dans les collectivites mal nourries. Con-

trairement a ce qui se passe chez le rat, la periode d'orga-nogenese durant laquelle prennent naissance les malfor-mations est relativement courte chez l'homme et, aucours de la longue periode foetale qui suit, les foetusanormaux peuvent etre elimines. I1 se peut qu'une partiedes avortements naturels chez les femmes mal nourriescorrespondent a l'elimination de ces fcetus non viables.Le rapport entre les trois deficits nutritionnels les plus

importants chez le jeune enfant - l'hypovitaminose A,l'anemie ferriprive et la carence proteique -, et le regimealimentaire de la mere durant la grossesse est discute.

L'etat nutritionnel de la mere se reflete dans la compo-sition chimique du lait, du moins sa teneur en vitamines.I1 sera interessant de determiner dans quelle mesure onpeut combler ces deficits par l'administration de vita-mines. Selon certaines observations, le retard de crois-sance durant la periode neonatale, dui a l'alimentationdeficitaire de la mere, serait irreversible. I1 y a lieud'examiner de quelle facon on pourrait ameliorer lesconditions de l'enfant au cours de cette periode decisive.Au cours de son expose, l'auteur signale les nombreux

points qui demandent encore a etre etudies, et suggeredes sujets de recherches.

REFERENCES

Athavale, V. B. (1960) Indian J. Child Hlth, 9, 75Bhavani Belavady (1959) Indian J. med. Res., 47, 217Bhavani Belavady & Gopalan, C. (1959) Indian J. med.

Res., 47, 234Bhavani Belavady & Gopalan, C. (1960) Indian J. med.

Res., 48, 518Dann, F. W. (1932) Biochem. J., 26, 1072Dann, F. W. (1934) Biochem. J., 28, 634Darby, W. F. (1950) J. Amer. med. Ass., 142, 1288Dawson, B. (1953) J. Obstet. Gynaec. Brit. Emp., 60, 80Dieckmann, W. J., (1952) The toxemias of pregnancy,2nd ed., St. Louis, Mosby

Ebbs, J. H., Tisdall, F. F. & Scott, W. A. (1941) J. Nutr.,22, 515

Gladstone, S. A. (1932) Amer. J. Dis. Child., 44, 81Gopalan, C. & Bhavani Belavady (1960) Fed. Proc., 20,

Suppl. 7, p. 177Gyorgy, P. (1960) Amer. J. clin. Nutr., 8, 344Jayalakshmi, V. T., Ramanathan, M. K. & Gopalan, C.

(1957) Indian J. med. Res., 45, 4Karmarkar, M. C., Chokshi, H. R., Shah, V. K. &Ramakrishnan, C. V. (1959) J. Pediat., 55, 481

Langley, F. A. (1951) Arch. Dis. Childh., 126, 64Lewis, J. H., Bodansky, 0. & Shapiro, L. M. (1943)Amer. J. Dis. Child., 66, 503

Lund, C. J. & Kimble, M. S. (1943) Amer. J. Obstet.Gynec., 46, 486

Lutz, P. & Platt, B. S. (1958) Proc. Nutr. Soc., 17, 3

Macy, I. G. (1949) Amer. J. Dis. Child., 78, 589Moore, T. (1957) Vitamin A, Amsterdam, Elsevier, p. 240Nelson, M. M. & Evans, H. M. (1946) J. Nutr., 31, 497Nelson, M. M. & Evans, H. M. (1953) J. Nutr., 51, 71Nelson, T. P. (1955) J. Obstet. Gynaec. Brit. Emp., 62, 49People's League of Health (1942) Lancet, 2, 10Scitchik, J. & Alper, C. (1956) Surg. Clin. N. Amer., 34,

1535Skurnik, L., Heikel, H. & Westerberg, T. U. (1944)

Z. Vitaminforsch., 15, 68Smith, C. H., Sisson, T. R. C., Floyd, W. N. & Siegel, S.

(1950) Pediatrics, 5, 799Steams, G. & McKinley, J. B. (1937) J. Nutr., 13, 143Thomas, J. W., Loosh, J. K. & William, J. P. (1947)

J. Anim. Sci., 6, 141Thomson, A. M. (1959) Brit. J. Nutr., 13, 190Venkatachalam, P. S., Kalpakam Shankar & Gopalan, C.

(1960) Indian J. med. Res., 48, 511Walker, D. M., Thomson, S. Y., Bartlett, S. & Kon, S. K.

(1949) The effect of diet during pregnancy on the vitaminA and carotene content of the colostrum of cows andheifers and on the reserves of the calf. In: Proceedingsof the 12th International Dairy Congress, Stockholm,p. 83

Warkany, J., Beaudry, P. H. & Hornstein, S. (1959)A.M.A. J. Dis. Child., 97, 274

Woo, T. T. & Chu, F. T. (1940) Chin. J. Physviol., 15, 83