Interacting with non-human animals (hereafter “animals”) has been associated with a range of well-being benefits among humans. Companion animal guardianship has been linked with improved physical and psychological outcomes, including lower blood pressure [1,2], reduced risk of cardiovascular disease and lower rates of mortality [3,4], reduced loneliness , and increased emotional support during mental health crisis . In fact, research has indicated that many people choose to keep companion animals for reasons associated with well-being, such as companionship, emotional support, and improved physical health [7,8]. Similarly, animal-assisted interventions (AAI) are initiated with the specific purpose of improving one or more aspects of human well-being; they include goal-oriented animal-assisted therapies delivered by healthcare professionals, and animal-assisted activities which are often volunteer-led and may lack specific treatment goals . These interventions have been used to support improvements in physical, psychological, and behavioural outcomes for a wide range of populations across the lifespan [10–14].
Despite these positive findings, research into the benefits of human-animal interaction (HAI) is far from conclusive. Some studies have shown no relationship, or a negative relationship, between keeping companion animals and physical or mental health outcomes [15–20]. Furthermore, as most research in this area is correlational it is difficult to determine causality; better health may increase the likelihood of adopting a companion animal, rather than the reverse , or health and companion animal guardianship may be linked by other factors, such as sociodemographic characteristics or health-related behaviours [19,22,23]. Similarly, while AAI are commonly perceived as beneficial, especially among those with a positive attitude towards companion animals, some authors suggest that these benefits have been overstated . In reality, research concerning these interventions is frequently anecdotal or descriptive in nature, with high levels of heterogeneity in factors such as the type of animal, the nature of the interaction, and the setting [25,26]. Methodological issues are also commonplace and include the absence of appropriate comparison groups, reliance on small samples, failure to randomise participants to conditions, and a lack of blinding for both participants and assessors [25–28]. It is therefore difficult to draw firm conclusions about the efficacy of these interventions [26–28].
These inconsistencies are further confounded by a lack of consensus about the mechanisms through which HAI may improve human well-being . Researchers have often referred to the biophilia hypothesis [29,30], which proposes that humans have an innate affiliation with other forms of life. This perspective suggests that because human evolution occurred almost exclusively in natural environments, people are predisposed to respond positively to aspects of nature that would have increased fitness in the ancestral environment, and negatively to those which would have decreased fitness . For example, people typically respond positively to natural landscapes providing sources of food, water or shelter, and negatively to animals which pose a threat, such as spiders or snakes . Although the biophilia hypothesis has been criticised for offering too broad a perspective and for lacking falsifiability , researchers have drawn on these ideas to develop theories with more explanatory power. For example, the biophilia-effect suggests that because the behaviour of animals is indicative of the presence or absence of threats in the environment, interaction with a calm or friendly animal may support human well-being by promoting relaxation and reducing physiological arousal [33,34].
Other popular explanations centre on the social support provided by companion animals. In the context of attachment theory [35–37] for example, humans are argued to form bonds with their companion animals which are comparable to those formed within close interpersonal relationships . This theory suggests that humans form strong emotional attachments with certain individuals, or “attachment figures”. These attachments are characterised by the presence of proximity seeking behaviours, distress at separation, and the provision of unique emotional support that cannot be replicated within other interpersonal relationships. Although attachment theory originally focused on the relationship between an infant and their primary caregiver (usually their mother), it was later expanded to incorporate the bonds which form in other close relationships, such as with siblings or romantic partners [39,40]. More recently, attachment theory has been applied to human-animal relationships, with findings suggesting that both the human and animal can serve as the attachment figure and provide feelings of comfort and safety during times of uncertainty or stress [20,38]. Furthermore, support provided by animals may be particularly effective, as it is unconditional and non-judgemental , and because physical touch–an important component of emotional support–is often discouraged with other humans but not with animals .
Alternatively, HAI may operate via distraction, whereby attention is diverted away from a perceived stressor to lessen the experience of negative mental states; this may be of most relevance in the context of AAI . Research has indicated that young children preferentially attend to images or videos of animals compared to non-living objects , and will choose to interact with real (but caged) animals over toys resembling those animals . Similarly, adults have been found to more rapidly identify changes in the location of living targets (animals and people), compared to inanimate objects . These findings suggest that animals may be particularly effective at attracting human attention. However, animals are not unique in being an effective source of distraction, and so similar benefits may be achieved through the use of alternative, and possibly more cost effective, stimuli [25,42].
This brief overview is by no means exhaustive and several other theories have been proposed [9,33,42]. Despite these divergent approaches however, one model provides a framework which may potentially incorporate some, or all, of the above discussed mechanisms. The biopsychosocial model  proposes that health is a continuum influenced by interacting biological, psychological and social factors; changes in one factor may influence the others and in turn impact health. For example, psychosocial stresses may lead to physiological responses including increased heart rate and blood pressure, or reduced immune function. Ultimately, these responses may have a negative effect on health, resulting in increased morbidity and mortality [47,48]. Equally however, some psychological and social factors may have a protective influence on health; higher levels of social support for example, have been linked to a reduced risk of cardiovascular disease . In the context of this model, there are numerous ways in which interaction with companion animals may impact human health. For example, owning a dog may lead to improved health through increased physical activity, while animals may provide social support either directly, or indirectly by facilitating social interactions with other individuals [49,50]. Conversely however, the grief associated with the loss of a companion animal may have a detrimental effect on human well-being . Thus, while the biopsychosocial model provides a potential framework for integrating multiple theories, it also highlights the likelihood that no one mechanism can account for the diverse effects of HAI. One area which warrants further consideration is whether the observed benefits are influenced by the type of animal involved in the interaction [26,51].
Multiple reviews and meta-analyses have noted that dogs are the animal most frequently involved in AAI (although other species such as horses are also commonly involved) [10,12,51]. Similarly, much research into companion animals has focused on those animals that can interact physically with humans, such as dogs and cats [3,52]. This type of interaction may not however, be suitable among all populations. For instance, people in rented accommodation are often restricted in the types of companion animal they may keep in their home, and physical interactions may be inappropriate for people with declining health or limited physical capacity . Similarly, dog-assisted (or similar) interventions often rely on volunteer services  and may require supervision of the client and animal to minimise risk, which can lead to infrequent and inconsistent exposure [53,54]. Issues may also arise where there is potential for aggression from the animal, where individuals have allergies, compromised immune systems, or phobias, or where contact with the animal could lead to accidental injury (e.g. scratches, falls) [51,55,56]. Animal welfare is also a concern, as some clients may behave aggressively or unpredictably towards the animal, or the animal may become stressed during the interaction [51,57]. Therefore, research into the effects of HAI with less physically interactive animals is needed to determine whether benefits may be experienced.
One form of HAI which has attracted relatively little investigation is the role of fish aquariums. Early research indicated a link between viewing fish in aquariums and benefits such as reduced blood pressure and increased relaxation [58–60], perhaps contributing to the widespread notion that aquariums are beneficial in healthcare settings . More recently, research has linked interaction with fish in aquariums to outcomes such as reduced anxiety , increased tolerance to pain , and improvements in nutritional intake and body weight among residents of specialised dementia units [64,65]. As with HAI research more broadly, the mechanisms underlying these benefits are unclear. Research with people who keep home aquaria has indicated that some individuals consider their fish to be a source of companionship, and feel an emotional bond with the animals ; this suggests social support and attachment may play a role in the beneficial effects of human-fish interaction. However, while research has shown the presence of attachment behaviours in other human-animal relationships, such as with dogs [66,67], it is not evident that fish exhibit behaviours such as proximity seeking or separation distress. Thus, while individuals may believe there to be an emotional bond between themselves and their fish, it is unclear whether this constitutes a true attachment bond as described by attachment theory [35–37]. Alternatively, watching fish swimming may simply be a source of distraction; this is supported by research which has shown positive physiological effects associated with viewing videos of animals, including fish .
An alternative perspective still comes from theories concerning the restorative value of nature. These theories suggest that exposure to unthreatening nature can help restore depleted cognitive resources, and support rapid emotional and physiological recovery from stressful events . Although most of this research has focused on natural or “green” landscapes, some studies have directly explored the role that encounters with wildlife play in human well-being. For example, research has suggested that many people feed wild birds because doing so brings them pleasure , while watching wild birds feeding is associated with increased relaxation and connectedness to nature . Similarly, improvements in self-reported control, happiness, and activity were observed among a sample of nursing home residents who were given the responsibility of caring for a bird feeder, while no changes were observed among residents who did not receive such an opportunity . Furthermore, research has suggested that for some individuals, wildlife encounters are a key motivation for visiting natural (coastal) environments , and are associated with a range of benefits to human psychological well-being . Given that the ways in which people interact with birds and other forms of wildlife are similar to the ways in which people interact with fish in aquariums (i.e. the interaction is largely visual), it may be that watching fish swimming promotes human well-being because this activity provides exposure to unthreatening nature, leading to restoration.
Irrespective of the mechanism, these findings suggest that human-fish interactions may be a viable alternative to more commonly researched forms of HAI. Furthermore, aquariums may overcome some of the issues associated with these forms of interaction. As a constant feature within the environment, fish aquariums are available to the client at any time and for as long as required, thus may provide greater flexibility in exposure than AAI which rely on visitation programmes . Even other types of resident animal cannot provide constant interaction, as this would be detrimental to their welfare . The monetary cost associated with installation and upkeep of a fish tank is also much smaller than that associated with other companion animals , although regular maintenance of the aquarium is needed, and requires an individual with knowledge of the necessary processes to ensure the welfare of the fish is not compromised. Aside from the person responsible for maintaining the fish tanks however, the passive nature of viewing fish in an aquarium means that even individuals with limited physical capacity are able to interact with the animals . There are no significant risks from aggression or allergies, and fewer risks associated with accidental injury due to the lack of physical contact with the animal (although possible injury could be sustained while installing or maintaining the tank, or if someone or something damages the tank, causing a break). While there is a small risk of bacterial infection associated with keeping home aquaria, this is rare and requires physical contact with the fish or water, so can be effectively minimised through careful hygiene practices .
Despite the potential benefits however, research in this area is limited, and to date only one review has sought to explore the potential benefits of fish aquariums to human health and well-being . However, this narrative review explored these benefits in the context of restorative environments and biodiversity, with a focus on the value of public aquariums; although there was reference to research conducted with home aquaria, this overview was not comprehensive. Furthermore, consideration should be given to the quality and strength of evidence when drawing conclusions from existing research findings; for this purpose, a systematic review of the literature is needed.