IJE Advance Access originally published online on February 29, 2008
International Journal of Epidemiology 2008 37(2):255-259; doi:10.1093/ije/dyn034
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary: The development of the Ounsteds’ theory of maternal constraint—a critical perspective
London School of Hygiene & Tropical Medicine, Keppel St., London WC1E 7HT, UK. E-mail: david.leon{at}lshtm.ac.uk
Accepted 4 February 2008
Margaret Ounsted conducted one of the earliest studies of the association between parental and offspring birth weights in humans published in the biomedical literature. From 19651 until shortly before her death in 1988, she published a series of papers that elaborated the notion of a maternal regulatory mechanism of fetal growth constraint. These have been continuously cited in the literature on inter-generational determinants of fetal growth.2–4 Over the past decade it has found a new and more general resonance in the context of the fetal/developmental origins of adult disease.5–9 What has attracted particular interest is the suggestion that the mechanism of in utero growth constraint is transmitted in a non-Mendelian fashion through the maternal line. Most tantalisingly, Ounsted hypothesized that the ‘set point’ of this mechanism is determined by the mother's own experience of constraint when she herself was in utero. These ideas were already evident in a paper published in Nature in 1966,10 although they are more explicitly articulated in the 1986 synoptic paper by Ounsted et al.11 reproduced in this issue of the IJE. It is the idea that a woman's own in utero experience may play a central role in the replication of maternal constraint across generations that strikes a chord with those working on fetal ‘programming’.
The concept of maternal constraint on fetal growth has a long history. Walton and Hammond's classic 1938 paper12 reported the results of crossing large Shire horses with small Shetland ponies. They found that offspring of the crosses delivered to Shire dams were heavier than that of pure Shetland ponies, but below that of pure Shire offspring. In contrast, the reciprocal cross-delivered to the Shetland dam was of the same weight at birth as the Shetland purebred foal. Thus, the Shetland mother was able to down-regulate the in utero growth of her foal sired by the much larger Shire horse, while the in utero environment provided by the larger Shire mother facilitated enhanced growth. As on average, crosses delivered to the Shetland and Shire mare would have the same genetic complement, they inferred the existence of a maternal regulatory mechanism linked to maternal size.
Although based on very small numbers, Walton and Hammond's paper is a model of elegant analysis and interpretation. Their conclusion was that the maternal regulation may be brought about by one or more of the following mechanisms: ‘(i) maternal regulation of fetal nutrition; (ii) maternal hormonal control; and (iii) cytoplasmic inheritance.’ Many of their key findings have been confirmed over the past decade in horses and cattle,13–15 and their paper continues to be widely cited both by animal researchers as well as those interested in human growth. Moreover, their idea of ‘cytoplasmic inheritance’ anticipates by half a century the relatively recent interest in non-Mendelian modes of inheritance through mitochondrial DNA16 and imprinted genes.17
In 1962, Margaret Ounsted moved from working as a medical officer with the blood transfusion service to the Nuffield division of obstetrics and gynaecology in Oxford.18 Within a short period of time she had set up a prospective study of intra-uterine growth retardation which continued to recruit subjects with minor modifications in design until 1975. First results were published in 19651 in a paper entitled ‘Maternal constraint on foetal growth in man’ which began with an account of the Shire x Shetland experiment. Her study was essentially a case–control design, although never referred to as such by Ounsted. The initial case series was 90 growth retarded or ‘small for dates’ (SFD) infants defined as those who at birth were more than 2 SD below the mean for their gestational age, using the 1958 British birth cohort as a reference.19 This SFD group were compared with a control series of 225 ‘unselected pregnant women’ seen at the ante-natal clinic before 20 weeks.
Ounsted was not impressed by the differences in maternal characteristics between her SFD and comparison groups. Gestational age, maternal age, maternal height and problems in pregnancy showed no statistically significant differences, and while smoking and social class were significant their effects were essentially dismissed as minor. Instead, what really struck Ounsted were the ‘gross and significant’ differences in the mean birth weights of live born siblings in the two groups. The SFD infants were much more likely to have growth impaired sibs than were the control series. Drawing on parallels with the Shire x Shetland crosses, Ounsted suggested that ‘growth restraint might be relatively constant in any given mother’.
A second publication quickly followed in 1966.20 This much less known paper extended the investigation to the birth weights of siblings of infants who were born ‘large for dates’ (LFD) i.e. >2 SD above the mean birth weight for their gestational age. The pattern observed was the ‘mirror image’ of that found among the sibs of probands of birth weight more than 2 SD below the mean: the sibs of the LFD babies were themselves large for their gestational age. In retrospect, one can see that these first two papers simply replicated the findings of studies published over 10 years earlier that showed the tendency for the birth weights of full-sibs to be correlated.21, 22 However, Ounsted did not appear to be aware of these prior studies in her first two papers.
A much more discursive and speculative discussion of these early results appeared later in 1966 in a Nature paper.10 This was the first of four papers Margaret Ounsted co-authored with her husband Christopher, a child psychiatrist ‘greatly influenced by Darwin’.23 It acknowledged the earlier work on consecutive births but crucially went further. It reported the preliminary results of what may be the first ever analysis of maternal birth weight in relation to offspring birth weight. Self-reported maternal birth weights of the SFD babies were found to be appreciably lower than the maternal birth weights of the control series of ‘unselected women’.
In their discussion, Ounsted and Ounsted10 highlighted work published in the first half of the 1950s on the genetics of birth weight. Morton had reported higher correlations of birth weight between maternal half-sibs than in paternal half-sibs,21 while Robson found higher birth weight correlations between first cousins on the maternal compared with the paternal side.22 Putting these earlier studies in the light of their own findings on maternal–offspring correlations in birth weight they concluded:
‘Those who have studied the patterns of intra-uterine growth within pedigrees have usually considered that the evidence shows the prepotent force of the maternal genotype. An alternative notion arises from the findings reported here. It seems possible that the degree to which the maternal regulator constrains intra-uterine growth is a function of growth constraint imposed on the mother when she herself was a foetus.’ (p. 996)
In retrospect, it is difficult to see what aspect of their data on mother–offspring correlations led them to put forward this programming-type hypothesis of intergenerational transmission. The Shire x Shetland cross experiment12 indicated that the in utero environment provided by the mother could have a major influence on fetal growth, but this did not provide direct evidence of inter-generational effects. More direct animal evidence was published some years after their 1966 Nature paper, when Stewart et al.24 reported on observations in rats undernourished for 10–12 generations. This showed impairments of fetal growth and adult size and function that took two generations of nutritional rehabilitation to remove.25 However, Ounsted and Ounsted did not have any human data that indicated that in utero constraint in humans leaves a mark on the reproductive performance of female offspring in the next generation. Moreover, subsequent investigations in humans of the impact of impaired fetal growth in one generation on that in the next have produced results that have been largely negative.26–29
Until 1975 Margaret Ounsted continued to recruit SFD and LFD infants and established a better-defined average for gestational age group (AFD) as a control, characterized by births not more than 1 SD in weight either side of the mean for gestational age. This final design, in which the SFD and LFD groups were compared with the AFD control group, has one feature that is rarely commented upon: the SFD and LFD groups are very extreme. Being either 2SD below or above mean birth weight for gestational age results in two groups each of which comprise 2% of births at either end of the full distribution. While this focuses on new borns who would have come to clinical attention, this design is far from optimal as a basis for generalising to the determinants of fetal growth in general. Moreover, Ounsted's SFD group is extreme even when contrasted with those who would today be considered small for gestational age. These are usually defined as those in the bottom 10% of the birth weight for gestational age distribution, not the 2% chosen by Ounsted.
As well as gathering birth weights and gestational ages of proband sibs, the pedigree data collected by Ounsted was extended to include self-reported birth weights of fathers as well as mothers, and uncles, aunts and first cousins. It is correlations among these different classes of relative that are the focus of the data presented in their 1986 synoptic paper11 republished here in the IJE, although much of this was already presented and discussed in a 1973 book by Ounsted and Ounsted.30
Several particular contrasts are emphasized in their discussion of the different types of familial birth weight associations. They claim that paternal size at birth is only related to offspring birth weight in the LFD group: ‘the father plays a negligible role when maternal constraint is prepotent.’11 However, it is clear from their own data (see Table 2 in republished paper) that the birth weight of fathers increases steadily from SFD, through AFD to LFD offspring, although not as steeply as the birth weight of mothers. Moreover, the existence of a clear contribution of paternal genotype to size at birth has subsequently been demonstrated in other studies.31–34 Most importantly, data analysed by Magnus34 provides no indication that the effect of paternal birth weight on offspring size varies according to birth weight of the mother. Nevertheless, these later studies do show maternal birth weight effects that are larger than the linear paternal effects, although as noted in another context, this sort of difference may be partly accounted for by misclassification due to unrecognized non-paternity.35
So where do we stand on the insights afforded by Ounsted? In their 1986 paper, Ounsted et al. reiterate three key ‘theoretical propositions’:
- In mammals, maternal systems constrain fetal growth rate to match that of the maternal strain.
- Maternal constraint is prepotent at the lower extreme; at the upper extreme relaxation of constraint allows other factors to take up more of the variance.
- The set point of the constraining mechanism is adjusted in utero in female fetuses.
The Ounsted's second proposition concerning the prepotent character of maternal constraint is more difficult to assess, mainly because the exact meaning of ‘prepotent’ is never precisely defined. One possible meaning is that non-maternal factors only exert an influence on fetal growth rate when maternal constraint is minimal or absent. However, as discussed above in the context of the association of paternal birth weight with offspring size at birth, there is little evidence of any interaction with maternal birth weight. To this extent, their second proposition does not appear to be supported, although it continues to be reiterated in the literature.38
Finally, turning to their third proposition, none of the human data available to Margaret Ounsted and her team provides support for the contention that a woman's own in utero experience of constraint determines the constraint she will impose on the fetal growth of the next generation. Nevertheless, this was something that the Ounsteds placed centre stage, arguing that such a mechanism could be regarded as having an evolutionary basis. In the final paragraph of the 1986 paper republished in this issue of the IJE, they discuss the adaptive significance of fetal growth rate being able to respond relatively rapidly to improvements in ambient nutritional environment. This same argument is made in even clearer terms by Margaret Ounsted39 in a 1971 paper:
‘It has been suggested that there is a maternal constraining mechanism which modulates fetal growth rate. Although the mechanism itself is probably specified genomically, pedigree data indicate that the actual set of the regulator may be determined by the degree of constraint imposed on the mother when she herself was a fetus. In populations chronically living under near famine conditions, the delivery of healthy but small infants at term is an advantage. Transmission of constraint by the method suggested would facilitate rapid changes in fetal growth rate, adapting different generations to variations in their living conditions.’(p. 526)Their argument seems to be that if in any one generation the nutritional environment improved, this would mitigate the intergenerationally transmitted tendency to constrain in utero growth, resulting in further diminished constraint in the succeeding generation. This foreshadows recent writing on developmental plasticity40 and predictive adaptive response41 that have been advanced as evolutionary frameworks for understanding the associations observed between growth in early life and risk of later coronary heart disease and diabetes.42 They share in common the proposition that the mother has an evolutionarily determined capacity to up- or down-regulate growth in response to information she herself receives when in utero. A thorough discussion of this more recent work goes beyond the scope of this commentary, although it should be noted that some of these ideas have not gone unchallenged.43
The Ounsteds believed that their evolutionary rationale was consistent with evidence that birth weight in populations did indeed respond rapidly to improvements in the environment. In their earlier work, they cited the apparent steep increase in birth weight in Japan following the Second World War as reported by Gruenwald.44 This spectacular improvement remains the exception. As discussed elsewhere45 there is little evidence of any substantial increases in mean population birth weight in high-income countries over the 20th century despite very substantial improvements in maternal nutritional status indicated by secular increases in adult height. Equally, the evidence that migrant groups, such as Indians coming to the UK, will see an increase in their mean birth weight in the second generation following migration is far from conclusive.46
In conclusion, Margaret Ounsted and colleagues made a real contribution to the literature. She was a pioneer in the study of parent–offspring birth weight correlations and brought to centre stage the distinctive role of maternal factors in determining fetal growth rate. However, her ideas about the prepotency of maternal constraint are of less value and the hypothesis that in humans the degree of constraint a mother exerts on her offspring's fetal growth is set by her own in utero experience is not supported by her own data or that published subsequently.
 |
Acknowledgements |
|---|
 Top Acknowledgements References |
|---|
Thanks to George Davey Smith, Michael Kramer and Kristina Stanfield who provided helpful comments on a draft of this commentary.
|
References |
|---|
|
TopAcknowledgementsReferences |
|---|
1 Ounsted M. Maternal constraint on foetal growth in man. Develop Med Child Neurol (1965) 7:479–91.[Medline]
2 Bakketeig LS, Hoffman HJ, Harley EE. The tendency to repeat gestational age and birth weight in successive births. Am J Obstet Gynecol (1979) 135:1086–103.[Web of Science][Medline]
3 Emanuel I. Invited commentary: an assessment of maternal intergenerational factors in pregnancy outcome. Am J Epidemiol (1997) 146:820–25.
4 Klebanoff MA. Paternal and maternal birthweights and the risk of infant preterm birth. Am J Obstet Gynecol (2008) 198:58–53.[Medline]
5 Barker DJ. The Wellcome Foundation Lecture, 1994. The fetal origins of adult disease. Proc Biol Sci (1995) 262:37–43.
6 Gluckman PD, Hanson MA, Beedle AS. Early life events and their consequences for later disease: a life history and evolutionary perspective. Am J Hum Biol (2007) 19:1–19.[CrossRef][Web of Science][Medline]
7 Kuzawa CW. Fetal origins of developmental plasticity: are fetal cues reliable predictors of future nutritional environments? Am J Hum Biol (2005) 17:5–21.[CrossRef][Web of Science][Medline]
8 Drake AJ, Walker BR. The intergenerational effects of fetal programming: non-genomic mechanisms for the inheritance of low birth weight and cardiovascular risk. J Endocrinol (2004) 180:1–16.[Abstract]
9 Gluckman PD, Hanson MA. Maternal constraint of fetal growth and its consequences. Semin Fetal Neonatal Med (2004) 9:419–25.[CrossRef][Medline]
10 Ounsted M, Ounsted C. Maternal regulation of intra-uterine growth. Nature (1966) 212:995–97.[CrossRef][Web of Science]
11 Ounsted M, Scott A, Ounsted C. Transmission through the female line of a mechanism constraining human fetal growth. Ann Hum Biol (1986) 13:143–151. Reprinted Int J Epidemiol 2008;37:245–50.[CrossRef][Web of Science][Medline]
12 Walton A, Hammond J. The maternal effects on growth and conformation in Shire horse-Shetland pony crosses. Proc R Soc Lond (1938) 125:311–35.[CrossRef]
13 Giussani DA, Forhead AJ, Gardner DS, Fletcher AJ, Allen WR, Fowden AL. Postnatal cardiovascular function after manipulation of fetal growth by embryo transfer in the horse. J Physiol (2003) 547(Pt 1):67–76.
14 Allen WR, Wilsher S, Turnbull C, Stewart F, Ousey J, Rossdale PD. Influence of maternal size on placental, fetal and postnatal growth in the horse. I. Development in utero. Reproduction (2002) 123:445–53.[Abstract]
15 Wilson ME, Biensen NJ, Youngs CR, Ford SP. Development of Meishan and Yorkshire littermate conceptuses in either a Meishan or Yorkshire uterine environment to day 90 of gestation and to term. Biol Reprod (1998) 58:905–10.
16 Chen XJ, Butow RA. The organization and inheritance of the mitochondrial genome. Nat Rev Genet (2005) 6:815–25.[CrossRef][Web of Science][Medline]
17 Constancia M, Kelsey G, Reik W. Resourceful imprinting. Nature (2004) 432:53–57.[CrossRef][Medline]
18 Moar VA. M Ounsted (obituary). Br Med J (1989) 298:45.[Web of Science]
19 Butler NR, Bonham DG. Perinatal Mortality. (1963) Edinburgh and London: E & S Livingstone.
20 Ounsted M. Unconstrained foetal growth in man. a preliminary note. Dev Med Child Neurol (1966) 8:3–8.[Medline]
21 Morton NE. The inheritance of human birth weight. Ann Hum Genet (1955) 20:125–34.[Web of Science][Medline]
22 Robson EB. Birth weight in cousins. Ann Hum Genet (1955) 19:262–68.[Medline]
23 Dennis J, Wooley H. C Ounsted (obituary). Br Med J (1993) 306:512.
24 Stewart RJ, Preece RF, Sheppard HG. Twelve generations of marginal protein deficiency. Br J Nutr (1975) 33:233–53.[CrossRef][Web of Science][Medline]
25 Stewart RJ, Sheppard H, Preece R, Waterlow JC. The effect of rehabilitation at different stages of development of rats marginally malnourished for ten to twelve generations. Br J Nutr (1980) 43:403–12.[CrossRef][Web of Science][Medline]
26 Glinianaia SV, Magnus P, Skjaerven R, Bakketeig LS. The relationship between maternal birthweight and gestational age in twins and singletons and those of their offspring in Norway. Paediatr Perinat Epidemiol (1997) 11:26–36.[CrossRef][Medline]
27 Lumey LH, Stein AD. Offspring birth weights after maternal intrauterine undernutrition: a comparison within sibships. Am J Epidemiol (1997) 146:810–19.
28 Stein AD, Lumey LH. The relationship between maternal and offspring birth weights after maternal prenatal famine exposure: the Dutch Famine Birth Cohort Study. Hum Biol (2000) 72:641–54.[Web of Science][Medline]
29 Hypponen E, Smith GD, Power C. Effects of grandmothers' smoking in pregnancy on birth weight: intergenerational cohort study. Br Med J (2003) 327:898.
30 Ounsted M, Ounsted C. On fetal growth rate. Its variations and consequences. In: Clinics in Developmental Medicine, No. 46. (1973) London: SIMP Heineman.
31 Klebanoff MA, Mednick BR, Schulsinger C, Secher NJ, Shiono PH. Father's effect on infant birth weight. Am J Obstet Gynecol (1998) 178:1022–26.[CrossRef][Web of Science][Medline]
32 Knight B, Shields BM, Turner M, Powell RJ, Yajnik CS, Hattersley AT. Evidence of genetic regulation of fetal longitudinal growth. Early Hum Dev (2005) 81:823–31.[CrossRef][Web of Science][Medline]
33 Little RE. Mother's and father's birthweight as predictors of infant birthweight. Paediatr Perinat Epidemiol (1987) 1:19–31.[Medline]
34 Magnus P, Gjessing HK, Skrondal A, Skjaerven R. Paternal contribution to birth weight. J Epidemiol Community Health (2001) 55:873–77.
35 Davey Smith G, Steer C, Leary S, Ness A. Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC). Arch Dis Child (2007) 92:876–80.
36 Pedersen J. Weight and length at birth of infants of diabetic mothers. Acta Endocrinol (1954) 16:330–42.[Medline]
37 Brooks AA, Johnson MR, Steer PJ, Pawson ME, Abdalla HI. Birth weight: nature or nurture? Early Hum Dev (1995) 42:29–35.[CrossRef][Web of Science][Medline]
38 Dunger DB, Petry CJ, Ong KK. Genetic variations and normal fetal growth. Horm Res (2006) 65(Suppl. 3):34–40.[CrossRef][Web of Science][Medline]
39 Ounsted M. Biological factors and fetal growth. Develop Med Child Neurol (1971) 13:524–28.[Medline]
40 Bateson P, Barker D, Clutton-Brock T, Deb D, D’Udine B, Foley RA. Developmental plasticity and human health. Nature (2004) 430:419–21.[CrossRef][Medline]
41 Gluckman PD, Hanson MA. Living with the past: evolution, development, and patterns of disease. Science (2004) 305:1733–36.
42 Barker DJ. The origins of the developmental origins theory. J Intern Med (2007) 261:412–17.[CrossRef][Medline]
43 Wells JC. Flaws in the theory of predictive adaptive responses. Trends Endocrinol Metab (2007) 18:331–37.[CrossRef][Medline]
44 Gruenwald P, Funakawa H, Mitaui S, Nishimura T, Takenchi S. Influence of environmental factors on foetal growth in man. Lancet (1967) i:1026–28.
45 Leon DA. Biological theories, evidence, and epidemiology. Int J Epidemiol (2004) 33:1167–71.
46 Harding S, Rosato MG, Cruickshank JK. Lack of change in birthweights of infants by generational status among Indian, Pakistani, Bangladeshi, Black Caribbean, and Black African mothers in a British cohort study. Int J Epidemiol (2004) 33:1279–85.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||