David Wastell and Sue White
The newly burgeoning field of epigenetics examines mechanisms which modify gene expression, mediating the interaction of the environment on what was assumed to be an immutable DNA blueprint. This invites a natural interest in the impact of adverse conditions, such as deprivation, or normatively deficient parenting, and aligns with the aspirations of the Developmental Origins of Health and Disease (DOHaD) paradigm:
“[A]n unhealthy lifestyle in prospective parents … passes greater risk of NCDs [non communicable diseases] to the next generation. This perpetuates cycles of poor health, reduced productivity and shorter life expectancy, trapping populations in a trough of low human capital from which they cannot easily escape.” (International DOHaD Society 2015: 1)
Under the spotlight are individual ‘choices’, including unhealthy eating, lack of exercise, smoking and alcohol use. In the same vein, The UK Houses of Parliament briefing on epigenetics and health stresses the need to equip practitioners to support behaviour change:
“Evidence from epigenetics research is [that] epigenetic changes are potentially modifiable through lifestyle and diet. Advice to pregnant women on behaviour change to avoid exposure to potentially harmful factors during early embryonic development is likely to be particularly important (Houses of Parliament 2013: 4)”.
In such a context, policy interest in all forms of early intervention from pre-conception onwards, is flourishing. However, there are limited studies on humans with the vast majority of the evidence coming from animal studies, principally rats. The work of Michael Meaney has been especially influential:
“Meaney and his group have shown how variation in maternal care can have long lasting effects… Offspring of mothers showing more maternal care are both less anxious and have a less pronounced corticosterone response to a new stressor. This [Meaney’s] group is also uncovering some of the epigenetic changes in the brain, altered methylation … which underlie this.” (Glover et al., 2010 p. 18)
Similar assertions can be found in the mainstream press:
“Research on rats by Prof Michael Meaney of McGill University, Montreal, and Frances Champagne … have identified changes in genes caused by the most basic psychological influence: maternal love. The 2004 study showed … that rat pups that had been repeatedly groomed by their mothers during the first week of life were subsequently better at coping with stressful situations than pups who received little or no contact.”
So, the seminal studies in the field are furnished by Micheal Meaney and co-workers (inter alia, 1985; 2001; Weaver et al 2004). Variations in the degree of maternal affection (“licking and grooming” LG and Arched Back Nursing ABN) apparently altered methylation patterns in the hippocampi of offspring; moreover, these epigenetic alterations could be reversed by cross-fostering with more attentive mothers. The most frequently cited study is Weaver et al. “Epigenetic programming by maternal behavior”, published in the prestigious journal Nature Neuroscience in 2004. Typing “Epigenetic” into Google Scholar, the paper comes up as the third most cited paper of all time, with 3588 citations (as of 4/11/15).
The study represents a landmark in the development of Meaney’s pioneering work on the stress response of rodents. It draws on a paper published thirty years previously (Meaney et al., 1985), which showed that laboratory rats which had been handled by experimenters and separated from their mothers for 15 minutes every day (placed in a plastic container away from the mother’s cage) showed 24% higher concentrations of Glucocortidoid Receptor (GR) in the hippocampus, a protein which places an important role in moderating the stress response. The paper does not explain why human handling and maternal separation during suckling is apparently so beneficial in developing a greater level of resilience.
That both handling (which involves maternal separation and isolation) and high nurturance seem to confer identical benefits in terms stress reactivity seems, for the lay reader, rather contradictory at first sight. Actually, there is a very clear link between handling and nurturance. A paper by Meaney and colleagues in 1997 (Liu et al, 1997) reports that handling has a dramatic effect on mother pup interactions: “Mothers of handled pups showed increased levels of licking and grooming and arched back nursing”. The average LG rate for the handled rats was 155 per observational period compared to 78 for the non-handled pups. The paper confirms that (human) handling is experienced as stressful by pups, it is “unsettling” leading to “increased vocalisation in pups, which in turn leads to more maternal care, including licking and grooming” (p. 1661). We now appreciate how handling makes rats more resilient. The mothers, presumably delighted to be reunited with their infants, licked, groomed and nursed with gusto.
Following the observational study of handling, the 1997 paper goes on to examine the impact of “naturally occurring” differences in maternal care on GR expression and stress responses in offspring. A group of mothers were examined; they were divided into two groups based on their LG-ABN scores, either above or below the mean level. On close inspection, the results are rather interesting from a purely behavioural of view. They suggest that a new phenotype has been appeared ex nihilo in study 2, the high LG-ABN “supermum. The rate of spontaneous maternal care for this group (an LG rate of 136) is nowhere to be seen in the first experiment, unless artificially produced by handling. So, there we have it, the missing link. The reason why handling confers the same benefits as high licking and grooming is because it engenders the same style of maternal care.
Let us look more closely. Three distinct mothering types are seen over the two experiments (Liu et al, 1997 and Weaver et al, 2004): high LG/ABN in response to a distressed infant, high LG/ABN where as far as we know no particular distress is involved, and low LG/ABN in a similarly non-stressed scenario. We also note, from the first experiment, that the latter was the norm for the laboratory mother; high LG/ABN was only produced by the intervention of the experimenters. A profound reversal has thus taken place: the normal has somehow become deficient, and the abnormal has become the new normal. This has the feel of a conjuring trick, but it is a legerdemain of the profoundest implications; entirely artificially, normal maternal behaviour has apparently been ‘pathologized’. Of course, there might have been differences in the way these behaviours were observed and coded, but we have been unable to find any such explanation and this, in itself, would surely signal a shift in normative understandings of ‘good enough’.
The seminal 2004 paper seems to present high LG-ABN mothering as a positive ideal. It is hard to read the paper in any other way than as a set-up to demonstrate this. No attempt is made, for instance, to examine whether the maternal style of the low LG/ABN mothers is better, especially in the artificial conditions of the laboratory where passivity could indeed confer adaptive benefits. No evidence is brought forward that offspring are in any way distressed by their “neglectful mothers”, so why treat it as negligent? But such critical reflection is not needed; Meaney (2001) has made his ideological position clear: “Cold, distant parent-child relationships are associated with a significantly increased risk of depression and anxiety in later life … warm, nurturing families tend to promote resistance to stress and to diminish vulnerability” (p. 1162). Of course, like the rest of us, experimenters are perfectly entitled to hold ideological positions, but these are, by definition, not morally neutral.
We may illustrate the direction of travel of such research by examining a further study, by Roth et al. (2009), explicitly focused on “dysfunctional mothering” designedly fashioned by the experimenters. A sample of rat mothers were deliberately subjected to stress immediately after birth: they were provided with limited nesting material in an unfamiliar environment and their rearing behaviour was then compared to unstressed mothers afforded with abundant nesting resources. The nurturing behaviour of the stressed mothers is explicitly described as “abusive” (Roth et al., p. 3): “pups were frequently stepped on, dropped during transport, dragged, actively rejected, and roughly handled” (p. 4). The maternal behaviour of these “maltreated-females” themselves displayed “significant amounts of abusive behavior towards their offspring” (p. 6) and in the “realm of normal maternal care” they frequently displayed “low posture nursing positions”. Their methylation levels were also described as aberrant; notably, a methylation inhibiting drug was found to reduce levels to ‘normal’.
Their conclusion is noteworthy. Commenting that “As epigenetic mechanisms continue to be linked with neuronal plasticity and psychiatric illnesses” they speculate pharmacological interventions such as “treatment with DNA demethylases or histone deacetylase inhibitors, might prove useful as therapeutic strategies for reversing persisting effects of early-life adversity (p. 8). Clearly, the single most obvious remedy which flows directly from the design of the experiment, is to provide mothers with adequate resources. Is a relevant solution to (inflicted) poor housing really the drug treatment of one’s offspring?
Despite the obvious caveats about extrapolating from rats to humans, especially rats in a completely artificial world, it is hard to read the animal studies as anything other than a proxy for human parenting. For some the leap is made with disturbing alacrity. There is a paradox here. The rats are not like humans and therefore can be subject to extreme stress, even torture, yet at the same time they are enough like humans to serve as “models”, and therefore can inform understandings of our own parent-infant bio-social relations. Anthropomorphism exists on one plane only.
A small but influential body of human studies of the effects of early adversity has followed in the wake of the rat laboratory work. There are robust studies on the impact of natural disaster, the Dutch Hunger Winter of 1944 and the Canadian Icestorm of 1998, but other studies on humans tend to examine the impact of what can only be described as varieties of maternal worrying. Let us look at an exemplar. Buss et al (2010) examine the impact of ‘pregnancy anxiety’ on brain morphology in 6-9 year old children. Maternal anxiety during pregnancy was measured using a pregnancy anxiety scale at 19, 25 and 31 weeks. The scale focused entirely on worries about pregnancy, health of the baby and fears about the delivery. These were not a clinical population with a diagnosis of mental illness.
From an original sample of 557 mothers, 35 agreed to MRI scans of their children. The children were screened and none had emotional, physical or behavioural difficulties; from this we assume that they were ‘normal’ children of ‘normally’ worried mothers. Pregnancy anxiety was not correlated with total grey matter volume but differences were found in a number of brain areas for women reporting higher anxiety in the first trimester. Despite the fact that these children are apparently normal, the authors go on to speculate that that reduced volume in areas of the prefrontal cortex might lead to delayed cognitive and motor development. Further speculations follow about how ‘higher concentration of stress hormones’ might cause further delays, except that for these children they apparently didn’t. The authors finally conclude that addressing women’s pregnancy related concerns should be ‘a major focus for public health initiatives’ (p149). This is of course an irony, since they have effectively fuelled maternal anxiety to some considerable degree by pointing to structural changes in the brains of (normal) children consequent upon fairly ordinary heightened anxiety in early pregnancy. It is hard to see this as progress. It also produces an ambition for the State that is both extraordinarily grandiose and likely to be futile: the eradication of worrying!
We ask, do we want to go in this direction without some reasoned debate and critical faculty being applied to claims being made in the name of human perfectibility?
Buss, Claudia, et al. (2010) “High pregnancy anxiety during mid-gestation is associated with decreased gray matter density in 6–9-year-old children.” Psychoneuroendocrinology 35(1), 141-153.
Glover, V, O’Connor, TG, O’Donnell, K, 2010, Prenatal stress and the programming of the HPA axis, Neuroscience & Biobehavioral Reviews, 35, 1, 17-22.
Liu, D., Diorio, J., Tannenbaum, B., Caldji, C. et al. (1997). Maternal care, hippocampal glucocorticoid receptors, and hypothalamic-pituitary-adrenal response to stress. Science, 277, 1659-1662.
Meaney, M. (2001). Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annual Review of Neuroscience, 242, 1161-1192.
Meaney, M., Aitken, D.H., Bodnoff, S. R., Iny, L. J., and Sapolsky, R.M. (1985). The effects of postnatal handing on the development of the glucocorticoid receptor systems and stress recovery in the rat. Progress in Neuro-psychopharmacology and biological psychiatry, 9, 731-34.
Roth, T.L., Lubin, F.D., Funk, A.J. and Sweatt, J.D. (2009). Lasting epigenetic influence of early life adversity on the BDNF gene. Biological Psychiatry, 65(9), 760-769.
Weaver, I.C.G, Cervoni, N., Champagne, F.A., D’Alessio, A.C., Sharma, S., Secjl, J.R., Dymov, S., Szyf, M. and Meaney, M. (2004). Epigenetic programming by maternal behaviour. Nature neuroscience, 7 (8) 847-854.
David Wastell is emeritus professor in Nottingham University Business School. Sue White is professor of social work in the School of Social and Public Policy at the University of Birmingham. The authors develop these and related arguments further in a forthcoming book to be published by Policy Press in 2017 with the working title Neuroscience, Genetics and Social Policy.