The majority of commentary on foetal pain has looked at the maturation of neural pathways to decide a lower age limit for foetal pain. This approach is sensible because there must be a minimal necessary neural development that makes pain possible. Very broadly, it is generally agreed that the minimal necessary neural pathways for pain are in place by 24 weeks gestation. Arguments remain, however, as to the possibility of foetal pain before or after 24 weeks. Some argue that the foetus can feel pain earlier than 24 weeks because pain can be supported by subcortical structures. Others argue that the foetus cannot feel pain at any stage because it is maintained in a state of sedation in the womb and lacks further neural and conceptual development necessary for pain. Much of this argument rests on the definition of terms such as ‘wakefulness’ and ‘pain’. If a behavioural and neural reaction to a noxious stimulus is considered sufficient for pain, then pain is possible from 24 weeks and probably much earlier. If a conceptual subjectivity is considered necessary for pain, however, then pain is not possible at any gestational age. Regardless of how pain is defined, it is clear that pain for conceptual beings is qualitatively different than pain for non-conceptual beings. It is therefore a mistake to draw an equivalence between foetal pain and pain in the older infant or adult.
During the past 10 years there has been increasing legislative interest in the possibility of foetal pain. In 2006, the US House of Representatives debated the Unborn Child Pain Awareness Act. The bill secured a majority but failed to obtain the two-thirds majority necessary to pass it as a law. Efforts at the state level have been more successful. At least 25 US states have deliberated on foetal pain legislation and at least eight (Alaska, Arkansas, Georgia, Oklahoma, South Dakota, South Louisiana, Texas and Wisconsin) now have legislation requiring that women seeking abortions be informed of the possibility of foetal pain.
Foetal pain has also been widely debated in Britain. The British Medical Research Council (MRC) and the Royal College of Obstetricians (RCOG) have both issued reports on this subject ; the issue of foetal pain was debated upon by the British Parliamentary Science and Technology Select Committee in 2008. To the author’s knowledge, however, no British or European legislation makes any direct reference to foetal pain.
Increasingly invasive surgical and other medical procedures performed in utero have also generated contemporary interest in foetal pain. Although rare, it is now possible to perform foetal surgery for conditions such as lower urinary tract obstruction, hydrothorax, cystic adenomatous malformation of the lung, congenital diaphragmatic hernia, spina bifida and large sacrococcygeal teratomas. More commonly, but still relatively rare, the foetus may be exposed to invasive interventions for the transfusion of donor red cells into the foetal intrahepatic umbilical vein or the peritoneal cavity. In addition, drainage of abnormal fluid collections, for example, a dilated bladder or hydrothorax, can be achieved by a single aspiration using a needle or the percutaneous insertion of an indwelling shunt into the amniotic cavity. Similarly, endoscopic placement of a balloon that is inflated in the foetal trachea can be used to improve outcome in cases of congenital diaphragmatic hernia. Finally, surgical abortion can obviously expose the foetus to intensely invasive procedures.
Resolving the question as to whether the foetus feels pain is not straightforward, most attempts at answering the question consider the neural pathways that are necessary for pain and ask when those pathways are present and functional in the foetus. This approach is useful because it is reasonable to assume that there is a necessary neural biology that renders pain experience possible. If the foetus lacks that necessary neural biology then it is reasonable to assume that it cannot feel pain. The approach is limited, however, by knowledge of the necessary neural biology. Although there is a general consensus that certain cortical structures are necessary for pain, legitimate arguments that cortical structures are not necessary continue to be raised. This article will first consider the neural pathways that might be necessary for pain and then go on to consider the limitations of deciding foetal pain based on the existence of such pathways.
The neuroanatomy of pain
A considerable amount is known about the biological structures involved in painful experience. Humans, and many other species, have neural structures that respond preferentially to noxious stimuli. Beginning in the periphery, there are nerve endings that preferentially transmit noxious information. They are the free nerve endings that arise mostly from the peripheral termination of A-delta and C fibres. The free nerve endings are polymodal and can respond to non-noxious and noxious temperatures or mechanical stimuli. When activity in A-delta and C fibres gives rise to pain or behaviour associated with pain then they are labelled as nociceptors. Fibres that only respond in the noxious range are labelled as nociceptive specific, while those that respond across the noxious and non-noxious range are labelled as wide dynamic range.
The primary afferent A-delta and C fibres terminate on neurons in the superficial dorsal horn of the spinal cord. Ascending projections to the thalamus originate from the most superficial layer, known as lamina I, and project contralaterally in the spinothalamic tract (STT). Intracellular recordings from lamina I neurons revealed neurons with seemingly modality-specific responses. One class of neurons were nociceptive specific, responsive only to noxious pinch, heat or both. Another class were thermoreceptive-specific, responding only to non-noxious cooling. A final class were polymodal, responding to heat, pinch and cooling. Provocatively, the existence of lamina I neurons, with specific responses and distinct morphology, motivated the suggestion that there are dedicated pathways for pain and temperature detection.
A series of neuroimaging studies have demonstrated consistent activation of several cerebral structures during pain. These structures include the primary and secondary somatosensory cortices and anterior cingulate, prefrontal and insular cortices. In combination, these structures are thought to coordinate defensive reactions and generate the sensory and unpleasant feelings associated with pain. Thus, a noxious stimulus sets in motion a train of biological events that acts to prevent further injury. People who lack nociceptors or have other biological deficits associated with the nociceptive system do not engage in defensive behaviours and do not report experiences of pain. Consequently, they suffer considerable, often life-threatening, injuries.
The first evidence for an intact nociceptive system in the foetus emerges at about 8 weeks gestational age (GA). At this stage, touching the perioral region will result in movement away, indicating the presence of sensory receptors and, at least, spinal or brainstem-mediated reflex action. Some claim that by 8 weeks GA, there are connections from the periphery and through the spinal cord to the thalamus ( http://www.abortionfacts.com/online_books/love_them_both/why_cant_we_love_them_both_14.asp ) but these claims are yet to receive any peer-reviewed verification.
At 8 weeks GA, the foetal brain is profoundly immature. There is no indication of maturation such as cortical sulcation and gyration or the appearance of a laminar structure in either the thalamus or the cortex. The external wall of the brain is about 1 mm thick, consisting of an inner and outer layer, but without a cortical plate from which the cortical layers will later develop. The cell density of the outer layer is significantly higher than that of a newborn or adult but contains large neurons that resemble those described in the older foetus. Beginning from about 9 weeks GA, there is thalamic fibre penetration directly into this outer layer that stimulates development and maturation of these large neurons. There is speculation that by 11 weeks GA these projections may be functional. The possibility of functional neurons from the periphery, into the thalamus and into the outer layer of the developing cortex places a lower time limit for foetal pain at around 11 weeks GA.
Between 12 and 18 weeks, the formation of the subplate begins and the first projections from the thalamus into the subplate appear. The subplate is a transient brain structure formed directly underneath the developing cortical plate. Neurons arrive in the subplate and are held for several weeks until the cortical plate becomes growth-permissive and facilitates neuronal invasion of the cortical plate. The relocation of neurons from the subplate to the cortical plate begins around 24 weeks GA and is extremely rapid from about 34 weeks. Afterwards the extracelluar matrix and other growth-related and guidance molecules disappear leading to the dissolution of the subplate.
Morphological features of maturity can be gradually observed from about 12 weeks GA. At 13 weeks, for example, a linear furrow or groove can be observed at the limit of the temporal lobe below and the insula, frontal and parietal lobes above. Around 15 weeks GA, this groove becomes part of the insular cortex, which is believed to be the first lateral cortical region to develop and is a key region involved in the experience of pain. The subplate has also been observed to thin in areas where cortical folding occurs, such as the parieto-occipital sulcus, and in the insula and cingulate gyrus at least from 20 weeks GA. It is currently uncertain whether this thinning is due to earlier maturation of these regions, and potentially earlier synaptic activity in the insula and cingulate cortex, which are both key areas in the experience of pain , or due to incidental morphological changes. Nevertheless, before 26 weeks GA, the foetal brain is largely smooth with only minor evidence of sulcation and gyration. Massive growth of the brain after 34 weeks rapidly results in the characteristic folds and surface features of the more mature brain.
By 24 weeks GA, substantial thalamocortical afferents have accumulated at the superficial edge of the subplate, which is the stepping-off point for axons growing towards their final cortical targets. Between 24 and 32 weeks there is substantial ingrowth of thalamocortical axons in the cortical plate of the frontal, somatosensory, visual and auditory cortex and formation of the first synapses in the deep cortical plate. Clear evidence of synaptic activity following auditory stimulation has been recorded from around 26 weeks in utero and somatosensory responses have been recorded in premature neonates of 25 weeks GA following a noxious heel lance.
By around 24 weeks GA, therefore, it can be assumed that noxious peripheral events cause a response in the primary sensory cortex indicating the presence of a spinothalamic connection. Long axonal tracts now course from the periphery and through the brain to the cortex. It is generally accepted that the necessary neural structures for pain are in place and are functional by 24 weeks GA. Thus, it is possible that measures to prevent pain might be appropriate during invasive procedures after 24 weeks GA.
Further developments
Recently, however, it has become clear that the pre- and post-birth environments are very different with potentially important consequences for painful experiences. It is also becoming clear that completion of the major pathways from the periphery to the cortex, at around 24 weeks GA, does not signal the end of cortical development but rather the beginning of a further maturational process.
In an extensive and important review, Mellor et al. described evidence suggesting that the foetus never enters a state of wakefulness in utero . This conclusion was based largely on observations of foetal lambs. Rigatto et al. , for example, directly observed an unanaesthetised sheep foetus, in utero , through a Plexiglas window, for a total of 5000 h without observing signs of wakefulness such as eyes opening or coordinated movement of the head. Mellor and colleagues suggest several factors explain this lack of wakefulness including the environment of the womb, which is warm, buoyant and cushioned, and the presence of a chemical environment that preserves a continuous sleep-like unconsciousness or sedation. The environment of the womb and the chemical suppression of wakefulness maintain unconsciousness and suppress higher cortical activation in the presence of intrusive external stimulation. The foetal response to hypoxia and asphyxia, for example, is characterised by apnoea, cessation of foetal body movements and a shift to a more quiescent Electroencephalographic (EEG) state indicative of unconsciousness.
A series of studies have also demonstrated that as spinothalamic pathways complete their connections with the cortex, from around 24 weeks GA, they increasingly stimulate intracortical pathways into development, which is the next major phase of neuronal maturation. This phase involves elaboration of dendrites and axons, formation and regression of synaptic connections and selective elimination of cell populations and corresponds to the cortical maturation described by Goldman-Rakic in primates and by Chugani in humans. McKinstry et al. illustrated the effects of this development using diffusion tensor imaging (DTI) in neonates born at 26 and 35 weeks GA. Neonates born at 26 weeks GA have a cortical cell structure dominated by radially orientated dendrites that inhibit the spread of water parallel to the surface of the brain. Consequently, DTI measures an elongated diffusion of water in the cortex. Neonates born at 35 weeks GA, in contrast, have a cortical cell structure with radially orientated and basal dendrites that inhibit the spread of water in all directions. Consequently, DTI measures a spherical diffusion of water in the cortex. This proliferation of cortical neurons and the overproduction of their arborisation and synaptic contacts begins prenatally, as illustrated by McKinstry et al. , but continues postnatally, along with synaptic elimination, pruning and programmed cell death.
The sedation of the foetus in utero , and the further development of the cortex, which only completes after birth, imply that it is only after birth, when the infant awakens and cortical development completes, that pain will be experienced. To put that slightly differently, it is seemingly reasonable, given the immature nature of the foetus and the apparent lack of wakeful activity, to reject the possibility of foetal pain on the grounds of pain being objectively impossible in utero .
Further developments
Recently, however, it has become clear that the pre- and post-birth environments are very different with potentially important consequences for painful experiences. It is also becoming clear that completion of the major pathways from the periphery to the cortex, at around 24 weeks GA, does not signal the end of cortical development but rather the beginning of a further maturational process.
In an extensive and important review, Mellor et al. described evidence suggesting that the foetus never enters a state of wakefulness in utero . This conclusion was based largely on observations of foetal lambs. Rigatto et al. , for example, directly observed an unanaesthetised sheep foetus, in utero , through a Plexiglas window, for a total of 5000 h without observing signs of wakefulness such as eyes opening or coordinated movement of the head. Mellor and colleagues suggest several factors explain this lack of wakefulness including the environment of the womb, which is warm, buoyant and cushioned, and the presence of a chemical environment that preserves a continuous sleep-like unconsciousness or sedation. The environment of the womb and the chemical suppression of wakefulness maintain unconsciousness and suppress higher cortical activation in the presence of intrusive external stimulation. The foetal response to hypoxia and asphyxia, for example, is characterised by apnoea, cessation of foetal body movements and a shift to a more quiescent Electroencephalographic (EEG) state indicative of unconsciousness.
A series of studies have also demonstrated that as spinothalamic pathways complete their connections with the cortex, from around 24 weeks GA, they increasingly stimulate intracortical pathways into development, which is the next major phase of neuronal maturation. This phase involves elaboration of dendrites and axons, formation and regression of synaptic connections and selective elimination of cell populations and corresponds to the cortical maturation described by Goldman-Rakic in primates and by Chugani in humans. McKinstry et al. illustrated the effects of this development using diffusion tensor imaging (DTI) in neonates born at 26 and 35 weeks GA. Neonates born at 26 weeks GA have a cortical cell structure dominated by radially orientated dendrites that inhibit the spread of water parallel to the surface of the brain. Consequently, DTI measures an elongated diffusion of water in the cortex. Neonates born at 35 weeks GA, in contrast, have a cortical cell structure with radially orientated and basal dendrites that inhibit the spread of water in all directions. Consequently, DTI measures a spherical diffusion of water in the cortex. This proliferation of cortical neurons and the overproduction of their arborisation and synaptic contacts begins prenatally, as illustrated by McKinstry et al. , but continues postnatally, along with synaptic elimination, pruning and programmed cell death.
The sedation of the foetus in utero , and the further development of the cortex, which only completes after birth, imply that it is only after birth, when the infant awakens and cortical development completes, that pain will be experienced. To put that slightly differently, it is seemingly reasonable, given the immature nature of the foetus and the apparent lack of wakeful activity, to reject the possibility of foetal pain on the grounds of pain being objectively impossible in utero .
The meaning of biological immaturity
The conclusion that the foetus cannot experience pain at any stage of development has met with considerable resistance. Opponents point out that the cortex might not be the only neural structure capable of processing pain. Subcortical structures, such as the brainstem, for example, respond to noxious stimulation. Subcortical responses occur in response to needling from 18 weeks GA. In addition, anencephalic infants, who typically lack almost the entire cortex, demonstrate a capacity to learn and show evidence of emotion , which supports the possibility of some kind of brainstem-mediated pain experience. An anencephalic foetus also withdraws from noxious stimulation, demonstrating that withdrawal is mediated at a subcortical level. Infants with significant neonatal neurological injury due to a parenchymal brain injury also respond to noxious stimulation with a pattern of biobehavioural and facial reactions similar to infants without brain injury.
If the foetus is sedated and asleep in the womb then it might be possible to dispense with all arguments based on neurobiology because it is typically assumed that a sleeping organism cannot feel. Mellor et al. propose that the foetus is unconscious based on the presence of powerful sedating chemicals in the womb (most notably adenosine), the presence of sleep-like EEG patterns observed in the lamb foetus that enter a more quiescent state during stress, and a lack of movement during stress. Unfortunately, these conditions do not guarantee or dictate a lack of wakefulness. Relaxation and sleep can be broken by noxious stimuli and although the EEG activity in the lamb foetus enters a state of relative quiescence during stress (such as caused by umbilical cord occlusion) there are clear indications of large spikes within that quiescence. Rather than EEG silence during stress, there is a shift from one EEG pattern to another and we have no direct means of assessing what either pattern means in terms of experience. It is also questionable whether the stress of umbilical cord occlusion is comparable to the nociceptive stress that will occur during surgery.
A series of studies with rats have also demonstrated that mammalian foetal behaviour is complex and organised. Foetal behaviour includes temporal rhythmicity, movement synchrony and motor coordination that are related to postnatal grooming, suckling and locomotor behaviour. The rat foetus responds vigorously to chemical stimulation, such as a lemon infusion, and preconditioning with other chemical stimulation, such as mint, modifies these responses. Human foetuses also detect and respond to chemical stimuli , move away when approached with a scalpel or needle during surgical procedures and show evidence of learning in utero . While these responses might occur during a state of sleep or sedation they are certainly not incompatible with wakefulness.
Associating EEG patterns during gestation with sleep-like states is also difficult because there is, in general, no clear cut-off between conscious and unconscious EEG activity. For the neonate, and presumably the foetus, sleep is accompanied by ongoing background EEG without obvious differences between wake and sleep. These observations of cortical activity during ‘sleep’ in the neonate raise the question of what ‘sleep’ and ‘wakefulness’ actually mean for the neonate or foetus.
A subcortical response to noxious stimulation and lack of uncertainty regarding the nature of foetal consciousness makes a definitive rejection of foetal pain difficult. The difficulty stems not just from a lack of neuroscientific knowledge but also from a lack of substance with regard to the nature of sensory experience. Without a satisfactory or comprehensive approach to the nature of sensory experience, terms such as ‘awake’ or ‘conscious’ or ‘pain’ and so forth take on a somewhat arbitrary character and can be attached to any apparently coherent neural behaviour. Certain patterns of EEG activity might or might not be associated with being awake; certain neuronal activity might or might not be associated with being conscious; and certain areas of the brain might or might not be necessary for pain. There is no means of deciding these issues without an adequate account of the subjective terms in use.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
