Can a fetus feel pain? This question is appropriate, as medical interventions during pregnancy,' fetal surgery,2 and possibly termination of pregnancy expose the fetus to noxious stimuli. To answer the question we need definitions of "feeling" and "pain." The International Association for the Study of Pain defines pain as "an unpleasant sensory and emotional experience associated with actual or potential tissue damage" and goes on to emphasise the importance of experience related to past injury. To feel something is defined as "to have the emotions excited, to experience a sensation." These definitions imply that the brain must achieve a certain level of neural functioning, as well as having prior experience, before pain can be understood.
Neural maturation ofany brain region includes so many stages and events that it is inadequate to use the presence or absence ofany one event as evidence offunctioning, but a critical factor must be the arrival of thalamocortical connections. These first penetrate the frontal cortical plate at 22-34 weeks,3 and their arrival correlates well with evoked potential studies, which show that the distinct component signalling the arrival of sensory impulses at a cortical level cannot be detected before 29 weeks.4
Role of sensory processing at lower levels
But is it appropriate in fetal and neonatal medicine to concentrate on the conscious perception of pain? Instead, should we not consider the neuronal activity that noxious stimulation produces in the pathways and nuclei concerned in sensory processing at lower levels of the central nervous system?
The development of the human spinal cord is a highly orchestrated process in which neurone and glial specific genes are induced in a manner closely similar to that in experimental animals.5 It becomes functional early in fetal life, and reflex movements to somatic stimuli can be evoked between 7.5 and 14 weeks while at the same time the fetus begins to move spontaneously in the absence of any obvious external stimulation. It is important to emphasise the reflex nature of these movements even though they may involve extensive body regions and therefore intersegmental and brain stem connections. They occur (in abnormal patterns) in anencephalic fetuses at 16-35 weeks even when only parts of the spinal cord are intact.6
Early reflexes are evoked by tactile stimulation or light pressure and are not necessarily nociceptive. In fact, small diameter afferent nerve fibres known to subserve nociception have not penetrated the fetal spinal cord at 19 weeks.7 However, by 26 weeks even very low birthweight infants show a clear and measurable flexion withdrawal reflex to noxious stimulation, suggesting that nociceptive afferent input to the spinal cord is present.8 Though these studies were carried out after birth, it is reasonable to assume that such responses would also occur in utero.
These cutaneous reflexes are exaggerated compared with those in adults, with lower thresholds and more synchronised and longer lasting reflex muscle contractions. Though providing important information on how the immature nervous system reacts to noxious stimulation, they cannot be equated with true pain experience, which must involve the cortex and develop postnatally along with memory, anxiety, and other cognitive brain functions. In other words, stronger reflexes are likely to reflect the absence of the normal inhibitory control or "dampening" influences that higher brain structures normally exert at more mature stages.9 Exaggerated reflex response to noxious stimulation may be protective to an organism which (by virtue of cortical immaturity or malfinction) is unable to perceive and organise a more directed response to the pain.
In addition -to spinal reflexes, infants delivered at 26-31 weeks show coordinated facial actions in response to heel prick.'0 In contrast with spinal cord reflex responses, younger gestational age is associated with less reactivity in facial expression to heel prick, suggesting that the youngest infants are less able to display more complex affective reactions to pain." In addition, from 23 weeks of gestation fetuses can mount a hormonal "stress" response when needles are inserted into the innervated intrahepatic vein.' Though noxious stimuli frequently produce a biochemical stress response, such a response does not necessarily indicate pain. The link between reported levels of pain and the hormonal stress response in adults is unpredictable and increasingly questioned.'2 What the fetal stress response shows is exactly the same as the behavioural studies-that noxious sensory stimulation can produce a clear reaction from the fetal nervous system.
There are great differences in neural structure and function between developing and adult pain pathways.'3 '4 One example is that fetal spinal cord sensory neurones which respond to noxious stimuli have larger receptive fields than in adults. Diffuse central connections and large dorsal horn receptive fields are likely to lead to poorer discrimination between noxious and non-noxious events and poorer spatial localisation by the fetus. Particularly important with respect to future analgesic therapy is the fact that neurotransmitter expression, receptor function, and distribution are all quite different from those in adults. Many immature receptor systems, including opiate receptors, are diffusely localised and have unusual patterns of expression," suggesting that opiate analgesic action in adults cannot be simply extrapolated to the fetus.
No "pain" but possibly altered neural development.
So can a fetus feel pain? Given the definitions of feeling and pain the answer must be no. But we cannot deny that the fetal nervous system mounts protective responses to tissue injury, starting just before the last trimester. The evidence for early exposure to noxious stimuli resulting in adverse effects on future neural development is increasing.l"20 In other words, noxious stimulation may not need to penetrate consciousness in order to substantially alter the course of sensory development. If a fetus or preterm infant is to survive and mature into an adult it is essential to consider this issue. As with the clear change that has occurred in the past seven years in pain management in neonates,2' it may now be pertinent to consider pain control in medical procedures for fetuses in the last trimester. The effects of trauma of any kind on the developing nervous system should be minimised as far as possible to avoid changing the course of normal development. More research is urgently needed on the developmental pharmacology of pain and analgesia.