Recognition of pain and distress

Discomfort and suffering

Discomfort is integral to animal life, and emotional correlates undoubtedly play a role in regulating the priorities among the various functions the organism must perform. Although we cannot know the subjective experiences associated with these emotions in other animals, we recognise expressions similar to ours, especially in species that are phylogenetically close to us.

We know that there are positive emotional states of pleasure, agreeability and satisfaction, which are the rewarding experiences associated with the successful execution of biological functions. Conversely, there are various ‘negative’ states of want, need and discomfort, reflecting a discrepancy between what is actually the case and the norm value of certain internal conditions or environmental relations. Such discrepancies are the motivational factors driving the behaviours that correct these discrepancies and reach and maintain the respective norm situations. Hunger, thirst, fatigue, the absence of a mate and the presence of a threat are prominent examples. These states respectively motivate the behaviours of feeding, drinking, sleeping, searching for mates and courting, and defence or avoidance. Within the natural boundaries of variation, such states can hardly be considered suffering. Although distress has many definitions, we use it in the sense of an aversive, negative state in which coping and adaptation processes fail to return an organism to physiological and or psychological homeostasis. ⁹ Discomfort turns into distress and suffering when such states persist at high intensities for long durations (from 1 to several days). This is likely to occur in higher animals, especially when the prospects of performing the appropriate corrective functions are unavailable and when the expectations of coping are low. Then, components of fear (a tendency to avoid specific objects, events and situations) and stressful anxiety (a more general and unfocused form of fearful arousal) make their contribution. The critical role of expectancies and the predictability and controllability of variables relevant to the animal in performing its functions has become evident since the pioneering experiments of Weiss.^10–12

This leads to the conclusion that various forms of activity, including group behaviour and social interactions, may be indicators of discomfort and distress. Appetitive behaviours associated with different functional systems may signal discomfort if:

The experiences of discomfort, need and hindrance must be regarded as biologically meaningful phenomena. Such experiences provide the impulse to execute behaviours that remove these states. If this does not ensue as a result of the behaviour in accordance with the animal’s expectations, the animal may switch to other behavioural modes. Pathological forms of behaviour, such as soothing stereotypies and directly damaging elements, may occur when animals are kept in a situation where conflicts cannot be solved, and discomfort turns into distress. If no available behaviour delivers the ‘expected’ result, the animal can also end up in a situation equivalent to depression or ‘learned helplessness’. ¹³ The animal loses the impulse to act and becomes passive, apathetic and listless. The switch to a ‘depressive’ attitude may be seen as a biologically meaningful reaction. Suppose coping attempts remain fruitless or even harmful. In that case, a strategy may be preferable in which the animal resigns from the situation, which is unmanageable, or suppresses all initiatives until the situation changes for the better. There is no point in wasting effort and running unnecessary risks. A similar apathetic attitude, characterised by depressed posture and unresponsiveness, can also be observed in cases of illness and chronic pain.

Pain

Pain represents a particular class of distress in both human and nonhuman animals. Pain can be defined as ‘An unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage’. ¹⁴ This concept can subsume a heterogeneous set of phenomena. It varies, for example, from the acute pain associated with certain forms of tissue damage. The aching of the muscles related to too much exercise, leading to the accumulation of lactic acid, to various forms of often chronic neural pain. In some of its forms, at least, a sensorineural substrate specifically tuned to nociception can be distinguished.

Pain has always been regarded as a biologically adaptive mechanism, the primary function of which is to warn the organism that (part of) it is under potentially damaging stress or that damage has actually occurred. Learning psychologists have emphasised an additional function. In the conditioning of behaviour, two kinds of effect influence the future occurrence of a behaviour: reward and punishment. The influence of the latter has been studied customarily by using mild electric shocks as a punishment. Here, pain acts as an unconditional stimulus, releasing withdrawal and defensive responses. At the same time, learning psychologists point out that associations are formed with the stimulus characteristics of the pain-provoking situation. These then act as conditional stimuli, releasing avoidance responses and fear towards that situation. Thus, the animal learns to recognise danger. However true, this role should not lead to a generalised statement that fear is a conditional response to pain. The ethological literature abounds with examples showing that flight and avoidance may also be evoked as unconditional responses, for instance, by sudden ‘startling’ stimuli or, more specifically, to certain sign stimuli, such as those indicating a possible predator. ¹⁵ Fear of predators cannot, as a rule, be a pain-conditioned response because few animals have a second chance after a painful confrontation with a predator.

An important distinction is made between primary and secondary pain.^16,17 This distinction is clearly recognisable in our own experience. The immediate sensation following a pain-inflicting stimulus is an acute, sharp and well-localised pain. This ‘primary pain’ is mediated by thick, fast-conducting myelinated Aδ-fibres. The primary pain often subsides soon, to be followed later by ‘secondary pain’. This is less localised and is generally of a chronic (or ‘tonic’) nature, sometimes pulsating. Even though potentially hurting severely, its quality is ‘diffuse’ and ‘aching’ rather than ‘sharp’. It is mediated by thin, slowly conducting, unmyelinated C-fibres. The relation between primary and secondary pain is complex. In the ‘gating theory’ of Melzack and Wall,^18–20 a mechanism is proposed for the short-term regulation of the transmission of incoming signals from nociceptive neurons via the spinal cord to the brain. The signals of the fast-conducting fibres, responsible for the initial acute pain, are, at the same time, thought to build up a self-inhibiting influence by clogging a spinal gate through which these signals are transmitted upwards. ²¹ This would explain that the primary pain abates soon after its sharp onset. When, later, the signals of the slow-conducting C fibres begin to arrive, this counteracts the inhibition, and the blockade of the gate is lifted. Thus, the nociceptive signals are transmitted once more to the brain. In addition to this short-term regulatory mechanism, there are other regulatory mechanisms of pain sensation, operating at more central levels under the influence of a diversity of motivational factors, for example, via enkephalins. ²² The transmission from the C-fibres appears to be very sensitive to opiate blockade; such an inhibitive influence has not been found for the fast Aδ-fibres. Bolles and Fanselow have argued that the different forms of pain, rather than being separate sensations, represent functionally differentiated aspects of an integral process of pain behaviour in response to tissue damage. ²³ They have distinguished three behavioural phases. The first two phases, the perceptive and defensive phases, are associated with the perception of primary pain. This triggers quick withdrawal responses in the organism or its affected body parts, as well as defensive responses aimed at warding off the pain-provoking agent. Reflex-like reactions of this kind can partly remain after brain connections have been severed. It seems that the more direct and reflex-like the reaction is, the more easily it can be dissociated from subjective experience, and the less trustworthy these are as objective measures of the pain experience. This also means that, although we can experience this pain, such experience is not necessary for the adaptive response. It also means that the occurrence of such reactions in animals does not necessarily imply perception and concurrent suffering. During the perceptual phase, the animal is also alerted to the pain-inducing agent and is conditioned to its stimulus characteristics. The third phase, the recuperative phase, is characterised by the onset of secondary pain. This pain triggers a motivational reorganisation of inhibiting activities that might interfere with recovery, allowing the animal to recuperate. The animal may look for a place to hide, rest and lick its wounds. Other functions are suspended. Posture and movements are depressed, just as in the human patient who is ill or suffers from chronic pain and displays a listless and apathetic attitude. Such ‘depression’ can be seen as a biologically adaptive response enabling the animal to be quiet and wait until it has recovered or (in the case of ‘helplessness’ after repeated failures) until the situation has improved. In the past, attention has been focused mainly on acute nociperception. Until recently, chronic pain in animals has received only limited attention. The LASA Working Party expressed their opinion that chronic pain or distress may often be more insidious, particularly in the early stages, than acute pain. ²⁴

According to Bolles and Fanselow, ²³ pain behaviour is best regarded as a separate motivational system that interacts competitively with other motivational systems. This is in line with ethological theory, which views behaviour as a hierarchical structure of motivational systems competing for dominance in expression. When a particular system achieves hegemony and its behavioural functions take place, this broadly inhibits other systems. This inhibition ensures that the programme of the activated function can be brought to completion, mainly without interruption. Just as the state of pain can suppress other motivations interfering with the recuperative process, the reverse can be factual (other motivations, when strong, can temporarily suppress the state of pain).

The recognition of pain

Pain recognition can occur in two circumstances. It may be triggered intentionally by applying noxious stimuli to evaluate the effects of specific experimental treatments (such as analgesic drugs or stressors) on pain perception. The methods target mainly primary and acute nociception rather than chronic pain. Typical stimuli include pinching, heat and electric shocks. Pain can also arise from both planned and accidental procedures. Gauging the animal’s level of suffering is vital in determining the acceptability of the methods. Accurate judgments about the extent of suffering and animal experiences should be based on the animal’s behavioural responses. It is important to recognise that different reactions may be expected depending on the stage of the pain behaviour process. This touches on an important prerequisite: the observer should be experienced. However, experience does not develop overnight. Reliable recognition of pain and suffering depends on thorough training and proper application. Standardised training, including inter-observer reliability measures, as described by Hawkins et al., ²⁵ is essential.

The reactions released by primary pain are generally withdrawal and protective responses. Additionally, the sudden onset of primary pain often elicits vocal responses in many species. The immediacy and conspicuousness of such responses make them easily recognisable as symptoms. It is essential to recognise that the threshold and intensity of responding can be influenced by a range of motivational factors, including stimuli related to predation risk, social conflict, territoriality and sex.

In the phase of recuperation, which follows when lasting damage has been inflicted, the responses are different and of a more heterogeneous nature:

In addition to the behavioural parameters discussed above, physiological features can also be important. We can distinguish between immediate and long-term physiological responses. Examples of the first are changes in pupillary dilation, cardiac rate, respiratory pattern, salivation, sweating, chromodacryorrhea, gastrointestinal motility, urination and defecation. Cardiac rate and body temperature will also increase due to handling, ³³ especially in small animals. An implanted telemetry device may be used to measure heart rate and/or body temperature^34,35; there are good reasons to believe that these reflect the animal’s experience. In some cases, such as certain internal afflictions in horses, these are the only signs visible. In other cases of suffering, long-term indicators may include impaired immune system functioning and increased levels of corticosteroids and catecholamines, which can be measured in both blood and urine. The measurement of some of these parameters requires invasive techniques, which can, in themselves, contribute to the disturbance. The stress involved in collecting blood samples for hormone measurements may modify the endocrine effects which one wants to measure. Blood sampling for corticosteroids should be performed within minutes in order to reflect corticosteroid levels before the stress of handling/restraint. ³⁶

Since steroid hormones are excreted in bile and faeces, ³⁷ and owing to the increased sensitivity of detection methods, data collection can be refined by measuring faecal glucocorticoid metabolites,^38,39 and hair cortisol, ⁴⁰ with minimal disturbance to the animal.

Catecholamines will increase immediately due to handling. In this instance, permanent catheterisation of the jugular or femoral vein can be the solution since it reduces the stress of handling during blood sampling. ⁴¹ These endocrine changes can inhibit gonadal activity and suppress reproductive functions. This can manifest itself behaviourally in inhibited sexual and nursing behaviours. Finally, growth may be impaired, and bodyweight may even decrease. Post-mortem parameters may provide clues about levels of discomfort, especially when these variables have been shown to correlate with clinical signs of discomfort in men and/or animals. They may include fatty depots, muscle volume, stomach ulcers, adrenal cortex size, lymphoid organs and fluid balance.

Another non-invasive method used in the study of animal welfare is infrared thermography. ⁴² Since cattle often perceive handling personnel as predators, they are highly motivated to conceal behavioural signs of weakness such as illness. ⁴³ Infrared thermography is very useful for the early detection of several important cattle diseases (e.g. bovine respiratory disease complex), although environmental factors must be considered. ⁴⁴ In larger animals, fear and other stress-related parameters have been measured successfully. ⁴⁵

Significance of pain signals

In nature, screams given as a response to primary pain can, just as fear screams, alert conspecifics to the source of danger. This might evoke help in defence from these conspecifics. Even if the alarm brings no direct benefit to the sender, the sender can nevertheless profit in terms of ‘inclusive fitness’. This can occur when there are relatives with a similar genetic background, who can then avoid the danger more effectively, for instance, by learning about its nature. ⁴⁶

Displays of pain and distress can also serve as signals of helplessness and need. Examples include the yelps and whines found in certain species, such as dogs and wolves. They are derived from the infantile repertoire, where they release parental care and protection. In some species, care has also been given to certain adult animals. It has even become a concern for group members other than parents. Examples include dolphins providing support to incapacitated pod members in danger of drowning, and canids where pack members bring and regurgitate food, not only to cubs but also to the adults staying with the cubs. In species such as these, signals of helplessness may release tolerance and even active support. ⁴⁷

Species differences in pain expression

Varied forms of emotional expression of secondary pain are expected to be especially characteristic of species that have evolved cooperation, sharing and communal brood care, such as socially cooperative carnivores, rather than in ungulates.

Social signals of primary pain are expected, especially in species where conspecifics can immediately and effectively adjust and avert the danger. Thus, pigs are very ‘touchy’ and react immediately and loudly when being squeezed. This is a highly adaptive response in a species where body contact is common, and piglets require strong signals when in danger of being trampled by heavy adults. In contrast, many other ungulates, such as antelopes, are comparatively mute, as audible signals of pain might attract predators.

Social facilitation and inhibition

Little is still known about contextual and environmental factors that might influence the expression of pain in animals. Concealment of pain behaviours may be expected in species and in conditions where their display might be a hazard. Signs of injury or sickness might attract and direct the attention of predators to easy prey; a limping individual in a herd of herbivores is an unmistakable signal. Also, in species with a social hierarchy, a display of weakness might tempt conspecifics to try and overthrow a dominant. Under such conditions of risk, other behavioural motivations might compete with the pain system and suppress it; this may consequently lead to increased pain thresholds. ⁴⁸ This reflects a trade-off in which specific short-term interests, for example, momentary safety, are secured at the price of a retarded recuperation. The evidence for such phenomena in animals is still largely anecdotal, and a systematic investigation is needed.

There is increasing evidence that social facilitation of pain expression, well-documented in the human species, also occurs in certain animal species. ⁴⁹ Such expression may be reinforced into ‘hypochondriac’ behaviour when an animal has learned that this brings social relief, tolerance or affection. In humans, conditioning processes may influence not only the expression of pain but also the thresholds of pain perception. Phenomena, reported especially for dogs, have been interpreted in this way, such as ‘sympathetic lameness’, limping, asthmatic behaviour and anorexia nervosa. Another interesting phenomenon is that of ‘empathy for pain’, which is defined as the mental capability to comprehend and respond to the feelings and emotions of other people accordingly. Empathy for pain means that following an observation of pain in a conspecific, the observer experiences similar sensations and feelings. This is a special kind of empathy, which is thought to exist in a wide range of animals, including those other than humans. ⁵⁰ Clearly, this provides a further complication when judging signals of pain and distress, especially when dealing with companion animals rather than experimental counterparts.

The impact of invasive experimental procedures on perceived stress and pain may be dependent on both physical and social environmental conditions. Both physical and socially enriched environments have been shown to affect the need for pain relief following painful experimental procedures. Pham et al. ⁵¹ showed that when, after surgery under anaesthesia, mice were given a choice to self-administer an analgesic available in one of their water bottles during 2 weeks postoperatively, socially enriched mice drank less from the analgesic-containing water than the non-enriched and socially deprived groups. Mice that underwent the operation self-administered more analgesics than mice that received only anaesthesia without an operation.

Grading the severity of pain and distress

A variety of schemes for scoring pain and distress in laboratory animals has been reported. Such assessments may be critical at three stages of experimental procedures: gaining approval from the animal ethics committee, during the experiment itself and in post-mortem examinations. The response to pain depends on factors such as age, sex, health status and the species and strain of the animal. The criteria used in pain evaluation are applied differently in various schemes. All current methods could be more satisfactory in that they are relatively subjective. Still, most workers agree that it is considerably better than doing nothing because the assessment of pain and distress implies explicit attention to the (potential) suffering of the animals used in procedures.^52–56

In general, discomfort can be assessed in either a qualitative or a more or less quantitative way. In both, the assessment of discomfort involves two steps: the collection of data, which can be regarded as an objective process, followed by the ‘translation’ of this data into a degree of discomfort, which is a subjective process. Morton and Griffiths, ⁵⁷ Beynen et al., ⁵⁸ FELASA, ²⁴ and the Disturbance Index used by Barclay et al., ⁵⁹ all attempted to score signs of pain and distress. FELASA identified components of severity and gave a numerical rating reflecting its potential range. ²⁴ Morton and Griffiths have correlated clinical signs and (severity of) pain and distress. ⁵⁷ In contrast, the Disturbance Index utilises changes in the number of movements made by a rat or mouse introduced into an unfamiliar cage to assess the severity of procedures. ⁵⁹

In Europe, severity is classified into four categories: mild, moderate, severe and non-recovery. Non-recovery refers to procedures performed under general anaesthesia without regaining consciousness. As a matter of course, severity is linked to the type of procedure. However, several factors may reduce or increase the severity, such as the age and sex of the animal, the type of manipulation, the animal’s habituation to the procedure, and the staff’s competence. A practical severity assessment requires several conditions, including all aspects of the culture of care within the institute (e.g. staff competence, use of standardised score sheets, clear lines of communication and well-defined responsibilities).

Development of behaviour recording systems such as LABORAS,^60,61 and Pheno World, ⁶² for the automatic registration of different behavioural elements of mice and rats, such as eating, drinking, grooming, climbing, resting and locomotion, can be used to assess behavioural changes as an indicator of discomfort in a less laborious way.

Today, artificial intelligence (AI) systems and machine-learning-based video analysis tools for continuous data collection and pattern detection are advancing rapidly. A nice example of such a system is DeepLabCut—a toolbox for markerless pose estimation of animals performing various behaviours.^63–66 The system has already been used successfully with rats, humans, different fish species, bacteria, leeches, flies, cheetahs, mouse whiskers and racehorses. Recently, the digital ventilated cage system has been developed for continuous home cage monitoring in small rodents.^67,68 Compared with human observations, these systems allow for continuous monitoring without fatigue, reduced observer effect, large-scale data collection and analysis, reduced operational costs over time, and more standardised and objective assessment of pain and distress.

Traditional observer-based home cage monitoring provides contextual judgment, and humans can detect unexpected behaviours that are not predefined in AI-based monitoring criteria, potentially leading to misinterpretation. In their systematic review, Kahnau et al. conclude that manual monitoring is declining relative to automated techniques but remains relevant. ⁶⁹

Using healthy animals housed in comfortable and stimulating environments will benefit both the animals and the experiment. Environmental enrichment can help create such an environment. This can be achieved by, e.g., giving opportunities to forage for food, social grouping, handling or training by humans, and structuring cages by providing, e.g. nesting material, nest boxes, climbing devices and shelters. ⁷⁰

Skillful experimenters will also minimise the animal’s discomfort by administering anaesthesia during painful procedures. Although anaesthetics and analgesics can affect an animal’s physiology and behaviour, and cause moderate discomfort during recovery, they are often mandatory for painful experiments or manipulations. In addition to pain relief, providing proper care such as warmth, soft bedding, palatable food and appropriate post-procedural analgesics is important.

Humane endpoints (HE) ⁷¹ should be considered in case of severe clinical signs, either by euthanising the animal humanely or by removing the animal from the experiment. In particular, when the clinical signs are an intrinsic part of the experimental procedure, it should be a conditio sine qua non to apply the HE as soon as the scientific endpoint is met. Other reasons to implement a HE are injuries resulting in severe and irreversible pain and distress (e.g. a rectal prolapse or bone fracture) or if the degree of suffering is beyond what has been anticipated (e.g. in case an infectious disease agent is more virulent than expected). HE parameters are generally linked to clinical signs; however, biomarkers from blood or urine (e.g. hormone levels or acute-phase proteins) can also be used. Sometimes, it is even possible to use pre-clinical parameters, such as antibody levels, instead of challenge procedures in vaccine potency studies, or to use non-invasive technologies, such as bioluminescence, to monitor tumour growth or metastasis. Applying HEs benefits the animal by eliminating or reducing unnecessary suffering. Additionally, researchers could benefit, as the experiment would be more valid, data would be less variable, and timely sample collection could be scheduled. The website https://humane-endpoints.info/ provides additional information on clinical signs and HEs.

Clinical signs that can quickly help determine humane endpoints are rapid weight loss (15–20% in a few days), an extended period of weight loss, prolonged diarrhoea (3 days), nasal discharge, coughing, neoplasia accounting for 10–20% body weight, self-induced trauma, icterus, central nervous system signs, severe ulceration or bleeding, drop in body temperature >4°C, prolonged inability to ambulate, laboured breathing and cyanosis.

A pilot study may help specify humane endpoints in particular experiments. The animals should be monitored at least twice daily if clinical abnormalities occur.