Persistence of pain in humans and other mammals

Abstract

Evolutionary models of chronic pain are relatively undeveloped, but mainly concern dysregulation of an efficient acute defence, or false alarm. Here, a third possibility, mismatch with the modern environment, is examined. In ancestral human and free-living animal environments, survival needs urge a return to activity during recovery, despite pain, but modern environments allow humans and domesticated animals prolonged inactivity after injury. This review uses the research literature to compare humans and other mammals, who share pain neurophysiology, on risk factors for pain persistence, behaviours associated with pain, and responses of conspecifics to behaviours. The mammal populations studied are mainly laboratory rodents in pain research, and farm and companion animals in veterinary research, with observations of captive and free-living primates. Beyond farm animals and rodent models, there is virtually no evidence of chronic pain in other mammals. Since evidence is sparse, it is hard to conclude that it does not occur, but its apparent absence is compatible with the mismatch hypothesis.

This article is part of the Theo Murphy meeting issue ‘Evolution of mechanisms and behaviour important for pain’.

1. Evolutionary models of pain

Evolutionary models have been largely overlooked in the study of pain. The widespread view is that while acute pain has clear survival value, chronic pain (outlasting healing) is an inevitable if rare malfunction of acute pain mechanisms, particularly peripheral and central sensitization, that ensure wound care and may heighten vigilance during healing. The neurophysiology of acute pain is extraordinarily conserved across animal phyla, vertebrate and invertebrate.

Pain in laboratory animals, mostly rodents, is usually studied over too short a timespan to provide data on chronic pain, and chronic pain reported in farm, companion and zoo animals in veterinary medicine is often associated with ongoing pathology. But if chronic pain outlasting healing is an inevitable if occasional by-product of acute pain, more accounts would be expected in animals that survive acute injury. It could simply be that observations are lacking, and that chronic pain does occur in wild animals. It is also possible that pain-related behaviour habituates and is only provoked in rare circumstances. But, if it is actually very rare, and chronic pain in humans derives some unique contribution from the human brain or environment, then the ‘inevitability’ model of chronic pain needs re-examination.

Chronic pain is often described as without adaptive function, in humans (e.g. [1]) and in other animals [2]. Two of the main evolutionary bases for vulnerability to disease proposed by Nesse & Stearns [3] apply to pain. Essential defence systems are subject to many false alarms (the smoke detector principle: [3]), given the balance of costs and benefits of action or inaction. An alternative explanation is mismatch of current conditions with evolved characteristics, since direction and intensity of natural selection can change with environmental shifts [4]. In wealthy societies, prolonged rest after onset of pain is possible, with no threat to survival; the same applies to domesticated and captive animals, provisioned and free from predation. This contrasts with ancestral conditions, and those for free-living animals, where survival may depend on activity despite pain.

This paper explores several questions: (i) do models of human chronic pain onset and persistence apply to other mammals; (ii) are behaviours associated with chronic pain in humans observed in other mammals; and (iii) what is the function of pain-related behaviours, particularly in relation to social responses?

2. Pain mechanisms

Mechanisms of pain are highly conserved across mammalian species. Nociceptor input to the spinal cord is modulated at spinal cord level and subsequent synapses by descending controls, excitatory and inhibitory [1,5]. Appraisal of threat in salience networks [6], contributions from memory, contextual and state considerations, and more are represented in descending pathways. Pain drives peripheral and central sensitization, changes in threshold and conductivity, changes in pre- and post-synaptic neurotransmitter release and uptake, and pruning and creation of connections in the brain. After injury, aversiveness of pain demands attention and action. After escape, heightened sensitivity and vigilance encourage wound care and avoidance of activities that could delay healing. So far, so adaptive: the function of pain is to promote escape, healing, recovery and learning, not to provide a quantitative representation of tissue damage.

Chronic pain, defined in humans as pain of more than three months, describes pain with and without identifiable cause: low back pain, joint pain from rheumatoid or osteoarthritis, neuropathic pains from traumatic injury or disease, and others; experience and impact are largely similar across types. Chronic pain has cognitive, mood and behavioural components, such as avoidance of activity believed to be threatening, thereby maintaining fears [7]. Theories now emphasize goal priorities [8,9], threat of pain and vigilance to internal sensations [10]. These models apply best to humans. The only model potentially applicable to other animals is that of classical conditioning, and is little explored: it proposes that a defensive response becomes conditioned to cues associated with pain that elicit fear and anticipation of pain [11].

3. Risk and protective factors

Given the common mechanisms for acute pain, it is useful to examine what predicts onset and persistence of chronic pain in humans, and whether the same risk factors are found in other mammals with chronic pain. The development of chronic pain is an active process [12]. Plastic changes throughout the system and reorganization in neural circuits subserving pain make for continued hypersensitivity to external stimuli and internal imbalances, for spontaneous pain discharges, and abnormal processing, increasingly in brain areas concerned with emotion rather than sensory input. Chronic pain may be driven by continued peripheral input, but can also be maintained by interactions with hormonal and immune system changes [13].

(a). Humans

Major risk factors for chronic pain in humans are characteristics such as female sex and greater age, the state of the individual in pain and the type of pain. For instance, in a prospective study of mixed chronic pains [14], the longer pain persisted, the less likely it was to remit: female sex and older age predicted both onset and persistence, and catastrophic thinking and sleep problems predicted more severe pain.

Chronic post-surgical pain follows traumatic injury of patients. Aside from peri-operative analgesia, one major predictor is the extent of nerve damage during surgery [15], driving abnormal nerve function. Pain before surgery is also a risk [15,16]; female sex and greater age tend to increase risk; while negative affect (such as fear of pain) consistently predicts worse post-operative pain [17] and development into chronic pain [18]. Anxiety may also predict persistence of pain and disability after traumatic injury, including amputation [16,19].

Chronic low back pain is more likely to develop in depressed individuals, with greater risk for more severe depression [20,21], and fear and anxiety are robustly related to subsequent disability [22]. Prospective fMRI study of the development of chronic pain confirms the importance of depressed affect before onset [23], the failure of descending inhibition and dysregulation of normally well-integrated cognitive, emotional and behavioural processes [24]. It may be that the reward value of pain relief is significantly disrupted [25,26].

Early-life stress, including pain (such as surgery or repeated blood sampling, common for preterm infants), affects the developing nervous system, producing changes in both pro- and anti-nociceptive mechanisms [27], against which maternal care may partly protect. Serious psychological stresses such as maternal death and institutional care in early years are also associated with higher risks of pain in adulthood, with some evidence of cumulative risk from multiple adversities [28].

Chronic pain may be prevented by the combination of exercise and education, and possibly by exercise alone [29,30]. Physical activity reduced risk of chronic pain in two large-scale prospective studies: one of over-50s for whom vigorous, but not moderate, physical activity protected against onset over 10 years [31]; another of female twins where genetic factors dominated covariance in onset of pain over 12 years [32], but environmental factors, including underactivity, predicted persistence of pain. In a major review, Sluka et al. [33] found fairly strong evidence for prevention of chronic pain by exercise, with some evidence for mitigation of existing pain (see also [34]).

(b). Other mammals

Are these predictors found in other mammals? Most veterinary and free-living mammal studies are uncontrolled, so factors such as sex and age have not been tested as risk factors for chronic pain, although likely [35]. Many rodent pain studies use young male rats [36], neglect affect and are short term. Parallels with the human literature are found for early-life stress, emotional processing of chronic pain and physical activity.

Rodent studies of early interrupted or impoverished maternal care show pups' later vulnerability to pain and hypersensitivity to various stimuli [27,28]. In a single study, ewes who had been tail-docked without analgesia when a few days old showed more pain at parturition than did controls [37]. Rodents, like humans, increasingly process pain in emotion circuits as pain becomes chronic [23,24].

Several rodent studies have established the protective effects of exercise against neuropathic pain [38], and the pain-reducing effects of exercise in neuropathic and low back pain [33,39,40]. The effects are associated with structural and functional changes in the nervous system and brain [39–42], in neuroimmune signalling [38] and in pro- and anti-inflammatory cytokine production [33,43]. Activity levels, among other environmental factors, produce epigenetic changes and, ultimately, changes in brain connectivity [44].

Studies of prevention and mitigation of pain by physical activity have only involved rodents, but survival in the wild requires that the injured animal balances reduced activity for recovery with the need to eat and drink sufficiently and not to become isolated. Continuing, even if at a lower rate, to be physically active, may attenuate peripheral and central sensitization processes. By contrast, companion, farm and other captive animals are protected from predation and other environmental hazards, and provided with food and drink; it is possible that by allowing inactivity, these conditions increase the likelihood of developing chronic pain.

4. Pain-related behaviour

Pain-related behaviours share features with sickness behaviours that conserve resources required for the immune response [45], under the influence on the brain of pro-inflammatory cytokines [46], and may enable others to avoid contamination or infection [47]. Healing and recovery may be sensitive to social cues [48]: the presence of familiar conspecifics or of recognized ‘healers’ conveys safety, allowing suppression of defence in favour of recovery, hence the placebo effect. Injury is common in contemporary hunter–gatherers [49], as is support for those who are injured. Steinkopf [50] describes symptoms both as defensive and as signals for help, which, when assured, renders symptoms redundant.

Particular attention is due to facial expression of pain, unique combinations of muscle actions that give a global impression of pain [51]. Although facial expression of emotion was initially conceptualized as a readout of internal state [52], it is better understood as a signal [53], making environmental and social context of critical importance.

Behaviour associated with pain on injury, across animals, was classified by Walters [54] in relation to selection pressures. Immediate and rapid behaviour includes arousal, active or passive defence, and memory formation; in the model of Bolles & Fanselow [55], this defensive stage is more associated with fear and inhibition of pain. By contrast, recuperation is associated with hyperalgesia, and in the Bolles and Fanselow model, with inhibition of fear. Although the Walters model describes recovery and healing with similar behaviours: protection by hiding, priming appropriate defences, keeping the wound clean, monitoring healing and conserving energy, it also emphasizes a high level of vigilance during recovery. The injured animal may have to choose between incompatible goals: of minimizing or relieving pain by immobility, thereby becoming more vulnerable to predation, or satisfying needs for food and water, at the cost of pain. Such decisions are made in the limbic system of the brain [19,23], but brain imaging (e.g. [56]) is underused.

(a). Humans

In humans, verbal report is assumed to be the most direct indicator, and used as a referent for interpreting other behaviour associated with pain (such as limping, guarding, taking analgesics and seeking social support). These behaviours are described mainly as attempts to avoid pain [7]. Behaviours may vary in function—protective versus communicative [57]—or be related to fear rather than to pain [58]. For comparison with other mammals, the focus here is on behaviours such as limping and guarding; changes in normal activities such as eating and social interaction, facial expression, and vocalization.

(b). Other mammals

Animal pain studies and observations present challenges of interpreting behaviour without verbal report, and in the context of cognitive, affective and social capacities. A cross-species approach to pain including motivational and affective responses [59] remains rare. A widely used definition of pain [2] resembles that for humans: ‘an aversive sensory and emotional experience’ representing damage or threat, promoting recovery and preventing recurrence.

(i). Rodents

Rodents are extensively used in studies of pain and analgesia [60], but few studies: use natural disease rather than inflicted injury [61]; assess spontaneous rather than evoked pain; include affective, cognitive and social responses; or follow for months [62,63]. Spontaneous behaviours, such as voluntary wheel-running or sleep changes, may sample an affective component of pain [60]; standard non-pain rodent tests of anxiety and depression are widely used, although they may be sensitive to analgesics [64]. Experimental paradigms that offer choice, such as conditioned place preference, or self-administered analgesia, provide proxies for self-report, and cumulative tracking of activity offers more ecologically valid outcomes [63]. Better recognition of the influence of environmental variables such as social conditions, handling or housing is required [36,61,65]; notably, the analgesic benefits of an enriched environment are becoming evident [66].

(ii). Companion and farm animals

Larger mammals may offer better opportunities than do rodents for translation of pain models to humans: they vary more in genotype and phenotype; affective and social variables are often observed; and living conditions may be closer to those of humans [65,67]. Much pain from farming practices and incidental disease is overlooked or underestimated [35,68,69]. Farm animals, young and adult, show specific behaviours that appear to be attempts to escape pain, such as changed posture and gait, and non-specific behaviours such as restlessness, withdrawal and facial expression, in acute (e.g. castration) and in some chronic pain situations [69,70], and in persistent disorders, such as mastitis [71] and lameness [72] in cattle. Some postural and activity changes observed in pain may also indicate mood change [73], and social context can suppress pain-related behaviour, as in post-operative horses [74]. For dogs, a scale estimating pain severity and interference is used across acute, chronic and cancer pain (e.g. [75]), although composite scales may obscure important differences in pain responses. Pain assessment in cats is less systematized, particularly in chronic pain; behaviour is very varied across gait, posture, rubbing, grooming, resting, appetite and social interactions [76,77].

(iii). Primates

Pain research using primates is relatively rare; welfare guidelines list behaviours such as reduced overall and social activity, appetite and changed posture and gait [78], but, in an endometriosis model of chronic pain in the marmoset, during extensive assessment, including cognitive tests and social behaviours, no behavioural changes were detected other than a slight decrement in social grooming, reversed by treatment [79]. This contrasts with human endometriosis pain-associated reduced activity and depressed mood.

Observations of parturition in various species of captive and free-living primates suggest pain during contractions through particular postures and facial expression. Injury is common in free-living primates from conflict, falls, predator attacks and environmental hazards such as snares. Post-injury, effective adaptation to feeding and locomotion seems to be the norm, with some reduced activity and avoidance of conflict [80,81], but few interpretations of behaviour in terms of pain. Such adaptation is not incompatible with pain; in species where rank determines access to food and mating, showing vulnerability could be seriously disadvantageous. Illness or disability is often associated with loss of rank, but the change may be small [82], and a study of the social consequences of disability in macaques observed neither discrimination against nor care for disabled individuals [83].

(iv). Other free-living mammals

There are few studies of pain in free-living mammals. Healing of injury appears common [84], even of long bone fractures or traumatic amputations, with survivors adapting effectively and with no unambiguous signs of chronic pain. For instance, a study of stags showed a lifetime rate of injury of 23%, 6% for permanent disabling injuries; both reduced reproductive success [85]. A study of roadkill deer found 11 of 24 had healed fractures, even of the pelvis, and one had lost her foreleg but carried a healthy fetus [86].

(c). Facial expression

Facial expression of emotion in mammals is highly conserved [87,88], but may only be detectable in chronic pain when acutely exacerbated. Pain ‘grimace’ scales exist for many mammalian species (e.g. [87]), despite the tenet that ‘prey animals’ suppress facial expression of pain (e.g. sheep [89]; horses [90]). Studies in the mouse suggest that emotional experience of pain is encoded in facial expression, as in humans [43], with variation according to sex and context [91]. The importance of facial expression of pain is that it is primarily communicative, which suggests a response of benefit to the signaller, so examination of social responses is warranted.

5. Social responses to conspecifics' pain-related behaviour

Various pain-related behaviours may communicate pain; behaviour as a signal transmits information at some cost to the signaller (deterring faking), and the responses of signal receivers provide fitness benefits [92]. Receivers may also be enabled to avoid pain. Responses may constitute instrumental or emotional help, no response, or exploitation of the vulnerable individual in pain, and may occur between carer and infant or young, between young or adults, between kin, familiars, competitors or rivals, and antagonists.

Signal and response might seem interdependent, but an experiment in artificial life showed that even when expression of pain/need disappeared, helping responses remained in the genotype [93]. Agents either expressed or suppressed ‘pain’ when injured randomly and so unable to ‘forage’ for energy; they either helped (donated energy to), ignored or exploited (stole energy from) other agents expressing pain. Agents with most energy ‘bred’ and passed on their characteristics. Over multiple baselines and many iterations, expression of pain continued when it cost little energy, injuries were infrequent, and there was no exploitation. Although an oversimplified model, it offers ways to explore interactions.

(a). Humans

Perception of another's pain activates the salience network of the brain of the observer [94]. Responses are often described in terms of empathy [95], but expression of pain does not necessarily elicit an empathic response [51], even by adults to children in their care [96]; and clinicians tend to discount patients' pain, believing their behaviour to be exaggerated [97,98].

(b). Non-human mammals

Responses to pain in a conspecific are rarely studied in non-human mammals, except rodents. The evolved capacity to tend to the needs of offspring could extend to other social attachments [94,95], and highly encephalized mammals with long-term social relationships are the most likely to show helping. Asian elephants comfort one another [99], but there are no accounts of pain. Ewes are attentive to lambs’ restlessness in pain from tail docking or castration [100], looking, sniffing and licking; it is unclear whether this mitigates a lamb's pain, although the presence of its twin may do so [101]. The only accounts of help to injured adults are in canines that routinely share food: among painted wolves (African wild dogs), wounded members are brought food and their wounds groomed [102].

Among adult primates, there is very little evidence of directed care, other than social grooming, with release of oxytocin and endorphins associated with reward and with analgesia [103]. Wounded individuals, in particular their wounds, may receive extra grooming that reduces risk of infection [80,81], and in captive and free-living populations, adults sometimes comfort and protect dying individuals (e.g. [104,105]); the contribution of pain recognition is unknown. Chimpanzees may wait for slow-moving disabled members, but there is no evidence of sustained care for injured members, who tend to adapt remarkably effectively even to major injuries [106]. By contrast, infants with disabilities, unable to cling or ride, are carried long past the usual age of independent mobility, with significant energy cost to the mother or other carrier [81,107].

Rodents have demonstrated surprising responsiveness to conspecifics' needs [91]. Mice orient to a littermate's pain and synchronize pain behaviour [108], mediated by facial expression, and sensitive to rank and stress in males. Female mice provide social analgesia to another in pain, choosing to spend time close to a ‘jailed’ mouse in pain at one end of the apparatus rather than to another ‘jailed’ mouse without pain at the other end, with the effect of reducing pain behaviour in the former [109]. In male mice, mild social threat produces hyperalgesia, but stronger threat analgesia [110]. Rats show emotional contagion and will work to terminate another's distress [111], but these studies do not involve pain. These findings are intriguing, because they occur between adults, in rodent species with different social structures and behaviours, and imply social analgesia from the unaffected individual's voluntarily staying close to the individual in pain. Social analgesia studies need replication in rats, but also in the companion and farm animals in which facial expression of pain has been identified.

6. Conclusion

Across mammals, pain neurophysiology is largely shared; risks and preventive factors are partly shared but much evidence is lacking; functions and trajectories of pain-related behaviours are poorly understood in both humans and other mammals. Pain-related behaviours in chronic pain may habituate or be suppressed, and tests such as self-administration of analgesia or avoidance of pain-related cues may be more sensitive than observation of spontaneous changes in movement, activity, and time budget. Where pain-related behaviours occur, they communicate pain, particularly to those close, and the possibility of social analgesia warrants investigation in other mammals. Pain needs to be studied in context, including broader lifestyle, to enable identification of conditions that might, as suggested by the mismatch hypothesis, contribute to chronic pain in humans and in domesticated animals.

(a). Chronic pain in non-human animals

It remains unclear whether chronic pain is a dysfunction of the pain system, a sustained false alarm, or an artefact of modern life, or some combination. The paucity of reports of chronic pain in free-living animals could be due to lack of observations, lack of behavioural indicators or rarity of survival with chronic pain. It would be perverse to propose that injured non-human animals do not feel pain as do humans, given the common anatomy and neurophysiology and plentiful evidence of recovery and healing. Peripheral and central sensitization are part of this phase, as rodent studies show. We know less of changes in the brain, or whether the contribution of anxiety, important in driving avoidance of activity and therefore increasing disability, is uniquely human.

In relation to the question of what causes acute pain to become chronic and to persist, some risks could be directly investigated in non-human mammals. The prediction of chronic pain by sex, age and extent of nerve damage is easier to investigate in various animals than pain and mood, but there appear to be parallel findings across human and non-human mammals in early-life stress and low levels of physical activity. Some living conditions would therefore be more conducive to development of chronic pain: a fixed home rather than nomadic lifestyle; sharing food rather than foraging individually. These would help evaluate the mismatch hypothesis of chronic pain.

(b). Behaviours associated with pain

Common observations of pain-related behaviours in humans and other mammals arise largely from studies of acute pain, where behaviour may be directed towards escaping pain, whereas in chronic pain, avoidance of exacerbation is predominant. Possible specific functions of behaviour are little discussed: stretching, writhing or rolling to try to relieve visceral pain; avoiding weightbearing for limb pain; scratching or biting at a superficial wound [73]. These behaviours might extinguish if pain persists, so would not occur in chronic pain. Generic behaviours, such as resting, conserve energy, but may also be an expression of helplessness in unremitting pain. The ultimate aim of behaviours is to restore homeostasis [24], but motivation and learning may be dysregulated in chronic pain [13]. These distinctions need to be addressed in future studies.

(c). Social context of pain-related behaviour

Young animals often seek and achieve contact with adults, often parents, although those adults' instrumental responses are few except in primates, who carry disabled young. Except for female mice, it is unclear whether adults of various mammalian species seek proximity when in pain, but the prevalence of facial expression of pain suggests that the social analgesia seen in mice may be far more widespread. The mouse studies [91,109] are replicable in other mammals. We do not know how a horse, rabbit, dog or sheep responds to another in pain; a familiar adult may be a safety signal [48,50], reducing the threat value of pain, and generating descending inhibition of pain.

Data accessibility

This article does not contain any additional data.

Competing interests

I declare I have no competing interests.

Funding

I received no funding for this study.