∗ , Heesook Son ∗ , Chia-Chun Tsai †
∗University of Maryland
†Yuanpei University, Department of Nursing
Health comprises the integration of psychological, physical, social, environmental, and spiritual aspects of an individual into a functional whole (Audy, 1971; Thomas et al., 2003, 2002; Sterling, 2003). Maximal health and wellness is life lived to its fullest. Individuals achieve optimal functioning along their personal continuums from minimal to optimal individual capacity. Individuals with personal, environmental, or physical limitations can achieve a high degree of health and wellness by living to their maximal capacities in a combination of these spheres (Audy, 1971). Healthy individuals live in harmony with themselves, others, and their environments.
Animal-assisted therapy (AAT) and animal-assisted activities (AAA) are two of several ways that animals can enhance or compromise individuals’ health. While this chapter focuses on physical indicators of health, it is important to remember the interconnections between the physical, social, and psychological components of health. Thus the psychological and social impact of friendly animals reported in Dr. Hart’s chapter will also impact the physical aspects discussed here. Psychosocial factors either promote health by moderating or promote disease by enhancing pathological processes (Audy, 1971). Psychosocial as well as physiological challenges play important roles in the pathogenesis of chronic disease (Thomas et al., 1997).
Animal-assisted activities refers to a general category of interventions without a common protocol. In general, AAA involves introduction of a companion animal to an individual who does not own that animal with the expectation that the introduced animal will provide short-term benefits to the individual at least while the animal is present. The benefits demonstrated from AAA are important components of benefits that can be derived from AAT. The impact of an animal on any one aspect of health will have effects on and affect other aspects. There is a great deal of variety in the implementation of AAA. It can involve the introduction of one or more animals of the same or different species to an individual in a private or group setting. The introduced animal(s) are accompanied by an individual responsible for the safe introduction and interaction of the pair.
The focus of most research addressing benefits of pet ownership or interaction with friendly animals stems from their potential to decrease loneliness and depression, reduce stress and anxiety, and provide a stimulus for exercise (Friedmann and Thomas, 1985). Stress, anxiety, and depression are associated with the hyperactivity of the sympatho-adrenal-medulla (SAM) system, the hypothalamic-pituitary-adrenal axis (HPA), and abnormal platelet reactivity (Musselman et al., 1998; Rozanski et al., 1999). Chronic stimulation of these responses increases likelihood of chronic disease morbidity and mortality (McEwen, 1998). SAM hyperactivity results in increased catecholamine release (Louis et al., 1975; Veith et al., 1994; Wyatt et al., 1971), reduced HR variability/increased sympathetic tone (Musselman et al., 1998), decreased myocardial perfusion and ventricular instability (Corbalan et al., 1974; DeSilva et al., 1978; Julius and Nesbitt, 1996; Skinner, 1985, 1981). In response to stress and depression, the HPA system initiates a series of neurohormonal responses and releases corticosteroids into the blood stream (Arato et al., 1986; Banki et al., 1992; Nemeroff et al., 1984; Pasic et al., 2003; Rozanski et al., 1999). The combination of stress, and anxiety and/or depression and the physiological components of these responses enhance the risk of cardiac mortality (Lampert et al., 2000; Rozanski et al., 1999). Excessive stress and psychological distress contribute to diseases of the skin, respiratory tract, as well as disruption in immune function and cardiovascular disease.
The presence of or interaction with friendly animals is conceptualized as a means of alleviating distress caused by loneliness and depression as well as decreasing physiological stress responses. The physiological outcomes studied as indicators of distress/stress include elevated blood pressure (BP), heart rate (HR), peripheral skin temperature and cortisol (Baun et al., 1991), in addition to risk factors for and mortality among patients with coronary heart disease.
Many applications of AAA are designed to benefit individuals by reducing stress and loneliness and inducing attention to and interaction with the outside world. Many of the studies that were designed to evaluate AAT would be evaluating AAA under more current nomenclature. In this chapter they will be termed AAA.
6.2 Stress-reducing health benefits from AAA
Within the past 15 years, several studies have directly addressed the impact of AAA on physiological indicators of health or stress/distress. The mixed results of the studies highlight the importance of considering the method of introduction of the animal and the way the AAAs are conducted in developing interventions to meet specific goals.
6.2.1 Individual AAA
Evidence for the success of an individual AAA at decreasing physiological indicators of stress is largely derived from studies using dogs with children and adults. In these studies, one animal interacts with each person in an individually oriented session.
The presence of a friendly dog was effective at reducing BPs and HRs of 2–6-year-old children undergoing simulations of routine physical examinations compared with the same children without a dog (Nagengast et al., 1997). Although the presence of a friendly dog was effective at reducing behavioral signs of distress in 2–6-year-old children undergoing actual physical examinations when compared with other children without a dog present, the physiological anti-arousal effects of the AAA were not replicated (Hansen et al., 1999).
Asking children to interact with a dog during a stressful activity was somewhat successful at reducing stress indicators. Havener et al. (2001) used AAA to reduce anxiety of 7–11-year-old children (N = 20) who were undergoing a dental surgical procedure. The children were encouraged to touch, pet, and talk to a dog that was lying beside them. There were no differences in peripheral skin temperature responses to the dental procedure between children with and those without a dog present during the procedure. A subgroup of 17 of the children indicated that they were stressed by coming to the dentist. In this group of stressed children, the dog’s presence moderated the stress response; there was less of a decrease in peripheral skin temperature between baseline and midpoint of the dental procedure in the children who had the dog present during the dental procedure.
Hospitalized patients also experience reduced stress when they interact with or watch companion animals. In a recent study, heart failure patients (n = 76) were randomly assigned to a 12-minute animal-assisted interaction (AAI) with a volunteer and therapy dog, a 12-minute interaction with a volunteer only, or usual care. The AAI group had significantly greater reductions in epinephrine and norepinephrine levels during and after the intervention compared with the volunteer group and significantly greater reductions in systolic pulmonary artery pressure and pulmonary capillary wedge pressure during and after AAI compared with the usual care group (Cole et al., 2007). Having an aquarium in their hospital rooms reduced stress levels of adult patients awaiting heart transplants (Cole and Gawlinski, 2000).
The stress-reducing effects of the presence of an animal extend to adult outpatients as well. Oncology patients who chose to have chemotherapy in a room with dogs present had significantly more improvement in oxygen saturation than those who did not have dogs present. Oxygen saturation actually decreased in the latter group (Orlandi et al., 2007).
Evidence from Holter monitoring of healthy older adults in their homes suggests cardiovascular benefits of the presence of pets. Four older adults who participated in a study of dog walking also agreed to have the dog spend time with them at their homes. During the 6 hours of home monitoring they spent two, 30-minute periods specifically interacting with the dog. Their cardiac function, as indicated by higher heart rate variability (HRV), was better when they were interacting with the dog than when the dog was not present (Motooka et al., 2006)
Attitudes toward an AAA animal influence its effectiveness. Playing with a friendly dog appeared to mute physiological arousal during play activities of pediatric cardiology inpatients who developed rapport with the dog during individual AAA. Heart rate, BP, oxygen saturation, and respiratory rate were recorded before, during and after 10–20 minutes of interacting with dogs by the patients and at least one parent. The increase in the child’s respiratory rate was negatively correlated with rapport with the dog. The better the rapport, the smaller the increase during interaction. Decreases in respiratory rate were most frequent during physical contact between the dog and the patient (Wu et al., 2002). Including the child’s physical exertion level in the analysis would strengthen the evidence.
Physiological stress indicator levels among health care professionals indicate that they also can benefit from AAA. Interacting with a therapy dog for 5 minutes led to decreased serum and salivary cortisol. The magnitude of the decreases in cortisol during 5 minutes of AAA was similar to decreases during 20 minutes of rest or 20 minutes of AAA (Barker et al., 2005).
While most of the studies of stress indicator reduction during AAA utilize a dog, other animals can provide similar benefits. Blood pressure and HR were lower during a moderately stressful activity after viewing videos of birds, primates, or fish than after control conditions, indicating the potential for many species to reduce stress responses (Wells, 2005).
6.2.2 Group AAA
The physiological effects of an AAA group and a child-life therapy group were similar for 70 children (mean age 9.9 years) hospitalized for an extended time, mostly for chronic diseases (Kaminski et al., 2002). In both groups, children were able to move around freely and choose activities of their choice with parents and/or staff and volunteers. Salivary cortisol, HR, and BP did not change significantly from before to after AAA or group child-life therapy or differ between the participants in the two therapies, despite observations of a more happy affect after AAA.
An aquarium in the dining room provided group AAA in a unique study of food intake among nursing home residents with dementia. The weight loss typically experienced in this population is due to failure to eat rather than changes in metabolic state. The introduction of an aquarium to the group dining room led to increased nutritional intake and weight gain. The aquarium held residents’ attention and encouraged them to spend more time eating (Edwards and Beck, 2002).
Individual AAA, in which a companion animal, usually a dog, was present during a stressful activity, was beneficial for reducing physiological indicators of distress in children and adults. The benefits of individual AAA were documented for the time the animals were present, but extension of the effects beyond the time of interaction was not documented. Group AAA did not improve physiological indicators of distress, but did encourage nursing home residents with dementia to remain in the dining room and eat.
AAA were demonstrated to be effective for reducing stress indicators and in situations where the individual’s response with and without the animal present were compared, but not when different individuals participated and did not participate in AAA. Inconsistencies of the results of investigations of the effectiveness of AAA highlight the importance of methodological design in evaluating its effectiveness. It is important to consider the appropriateness of the specific implementation of AAA to obtain the desired outcome. Schwartz and Patronek (2002) provide excellent insight into many of the methodological issues to be considered in design and interpretation of studies assessing the anxiety-reducing effects of AAA.
A larger and ever-increasing body of research provides a theoretical basis for positive impact of AAA and AAT on human health. Some of the research relies on epidemiological research methods that study groups of people in their natural environments while other research relies on experiment data from studies over short durations in laboratory or home conditions. Two intervention studies examined the physiological impact of acquiring pets.
6.3 Epidemiological evidence for health benefits
Epidemiological methods allow non-manipulative investigation of the association between specific characteristics or exposures and health outcomes by examining large groups of subjects in their natural settings. Single epidemiological studies provide evidence of association but are not conclusive with respect to causation. The combined evidence from several epidemiological studies provides strong support for causation of health outcomes, usually mortality or morbidity.
The integrative aspect of the various components of health is demonstrated by the combined contributions of social, psychological, environmental, and physical factors to chronic diseases. Coronary heart disease was among the first chronic diseases for which the contribution of social and psychological factors was demonstrated (Jenkins, 1976b,a). Pets were conceptualized as a contributor to the social aspect of health. The cardiovascular system was a logical starting point for evaluating the possible effects of owning pets on human health (Friedmann et al., 1980).
Several case control studies demonstrate the association of owning a pet with cardiovascular health. In the first study of this type, pet ownership was associated with survival among patients who were hospitalized for heart attacks, myocardial infarctions, or severe chest pain, angina pectoris (Friedmann et al., 1980). Only 5.7% of the 53 pet owners compared with 28.2% of the 39 patients who did not own pets died within one year of discharge from a coronary care unit. The relationship of pet ownership to improved survival was independent of the severity of the cardiovascular disease. That is, among people with equally severe disease, pet owners were less likely to die than non-owners. Owning a pet did not appear to substitute for other forms of social support such as being married or living with others. This study was replicated and extended to a larger number of subjects with improved measures of cardiovascular physiology and psychosocial status (Friedmann and Thomas, 1995). Among 369 patients who had experienced myocardial infarctions and had ventricular arrhythmias, followed by life threatening irregular heartbeats, both owning pets and having more support from other people tended to predict one-year survival. As in the previous study, the association of pet ownership with survival could not be explained by differences in the severity of the illness, psychological or social status, or demographic characteristics between those patients who owned pets and those who did not.
The possibility that some species of animals might provide distinct benefits to their owners while others might not, gained limited support from epidemiological evidence. In Friedmann and Thomas’s (1995) study, dog owners were approximately 8.6 times more likely to be alive in one year as those who did not own dogs. The effect of dog ownership on survival did not depend on the amount of social support or the severity of the cardiovascular disease. In contrast, cat owners were more likely to die than people who did not own cats. The relationship of cat ownership to survival was confounded by the effect of social support, which was low among cat owners and among those who died, and by the over-representation among cat owners of women, who were almost twice as likely to die as men. A subsequent study of 6-month survival among 454 patients who were admitted to a hospital for myocardial infarction also suggested that cat ownership might have different health impacts than dog ownership (Rajack, 1997). Cat owners were more likely to be readmitted for further cardiac problems or angina than people who did not own pets. However, in contrast to the previous studies, pet ownership was not related to 6-month survival, or to other indicators of health. The one difference between dog and cat owners’ cardiovascular health must be interpreted cautiously; one significant difference among many comparisons raises the possibility of a chance effect.
Pet ownership may protect people from developing coronary heart disease or slow its progression in addition to influencing the survival of individuals who have experienced myocardial infarctions. Several cross-sectional studies and one longitudinal descriptive epidemiological study addressed the differences between pet owners and non-owners in health indicators. Among 5,741 people attending a screening clinic in Melbourne, Australia, risk factors for coronary heart disease were significantly greater among the 4,957 pet non-owners than among the 784 pet owners. For men, plasma levels of cholesterol and triglycerides, and systolic BP were higher among pet non-owners than pet owners. For women, differences in risk factors between pet owners and non-owners occurred only for those women who are most susceptible to coronary heart disease, women in the menopausal and post-menopausal age groups (Anderson et al., 1992). A study of senior citizens (n = 127) also indicated that pet owners have lower serum triglyceride levels than non-owners (Dembicki and Anderson, 1996). In contrast, a random sample of 5,079 adults from Canberra and Queanbeyan, New South Wales, Australia, interviewed in 2000 and 2001—approximately 57% of whom were pet owners—revealed no significant reduction in cardiovascular risk factors for pet owners or in use of health services (Jorm et al., 1997; Parslow and Jorm, 2003). In fact, pet owners had higher diastolic BP after controlling for age, sex, and education than those without pets (Parslow and Jorm, 2003). When the older group—those 60 to 64 years old—was examined separately, there was no evidence for a health benefit (Parslow et al., 2005). Differences in patterns of pet ownership may be responsible for the apparent discrepancies. Pet ownership was considerably more common in the New South Wales survey than in Melbourne survey. Types of pets were not evaluated in the New South Wales survey.
In several surveys, owning a pet was related to proxies for physiological health such as medical visits, number of health problems, or functional status. In large representative sample surveys of the populations of Germany and Australia, after taking into account demographic predictors of health status, people who owned pets made fewer medical visits than those who did not (Headey et al., 2002). In the USA, pet owners (n = 345) among the 938 Medicare enrollees in an HMO reported fewer medical visits including both fewer total doctor contacts and fewer respondent-initiated medical contacts over a one-year period than non-owners (Siegel, 1990). Further analyses of the data indicated that pet ownership was a significant moderator of the impact of psychological distress on doctor contacts, independent of the effects of health status, depressed mood, and other demographic factors. For individuals who did not own pets, psychosocial distress, as assessed by stressful life events, was directly correlated with doctor contacts; the higher the stress level, the more contacts. However, for pet owners increased stress levels did not predict more physician contacts. There was also evidence for differences in the effects of dogs and other pets on health as assessed by health behavior. For individuals who did not own dogs, doctor contacts increased as life events increased. In contrast, among dog owners, life events were unrelated to respondent initiated doctor contacts.
A Canadian longitudinal telephone survey of adults 65 years and older also supported a positive impact of pet ownership on respondent’s (n = 995) ability to complete activities of daily living at study entry and one year later. After controlling for physical activity, age, and living situation, the ability to complete activities of daily living decreased more in one year for people who did not own pets than for people who kept them (Raina et al., 1999).
One of the major questions arising in studies finding an association of pet ownership with health status is whether the data are due to people with better health status choosing to own or interact with pets. If so, the better health could predate the pet exposure. Thus, the causal relationship would begin with better health status and end with pet ownership. The question of which came first, pet ownership or better health was addressed in a novel longitudinal study. A cohort of 343 people entered into a population study in 1921 was asked in 1977 about their history of playing with pets; this was then followed for 15 years. In 1977, there was no relationship of pet-related behavior to long-term survival. This was true even when looking separately at individuals with low social support (Tucker et al., 1995). These data do not support the supposition that better health predates or causes more interaction with pets.
Introduction of an animal into a living situation can lead to improved health status. Adopting a pet was associated with improved health status for the adopters (Serpell, 1991). People who adopted dogs or cats from an animal shelter (n = 71) experienced significant reductions in minor health problems including headaches, hay fever, painful joints, one month after adopting the pet. Dog adopters (n = 47) maintained the decrease in minor health problems over the 10-month duration of the study; cat owners did not. In this study, dog owners both appeared to walk slightly more at baseline and reported increased frequency and duration of walking at 10 months. This suggests the possibility that adoption of a cat could have encouraged the owner to spend additional time at home and thus forego walks. The physiological benefits associated with acquiring a dog could have been the result of increased physical activity engendered in walking the animal. In fact, the absence of long-term benefits for cat owners supports this possibility. However, those who adopted dogs already tended to walk more at baseline than those who adopted other animals and the control group. The differences in walking may have been a reflection of other differences in lifestyle and availability of time for walking and caring for a dog. Differences in the health experience of those who adopted dogs and cats also may be confounded by differences in stressful life events. Introducing pets and plants into nursing homes similarly was associated with improvement in minor health problems as evidenced by reduction in amount spent on medications (Montague, 1995).
Dog and cat ownership might have different associations with health status as evidenced by two case control studies (Friedmann and Thomas, 1995; Rajack, 1997) and two longitudinal studies (Serpell, 1991; Siegel, 1990). Too few pet owners own only other species in these studies to begin to explore differences in health among them. The mechanisms for differences in health status of dog and cat owners as well as which aspects of health might be affected by each species remains to be evaluated. Apparent difference in health benefits of dog and cat ownership, with the exception of exercise-related benefits, many be reflections of differences between people who choose to own different species (Serpell, 1991).
6.4 Experimental or quasi-experimental research
In an attempt to understand how, from a physiological perspective, pets provide the benefits detailed above, a number of researchers have investigated the short-term effects of companion animals on people. These short-term effects, measured on the time scale of minutes rather than months or years, may be the bases for the long-term effects demonstrated in epidemiological studies as well as for other more subtle effects of pet ownership.
The vast majority of the studies of the effect of animals on human physiology utilize experimental techniques in which the physiological effect of an image of an animal or an animal stimulus is measured. Although the epidemiological studies cited above include pets of all types, a majority of the studies of the short-term impact of animals on human physiology are limited to the effects of dogs. This is largely a matter of convenience because dogs are kept as pets so frequently and they are easy to handle. In the research investigating the short-term stress-reducing effects of animals, two types of potential health benefits were investigated: direct effects on physiological indicators of stress and stress moderating or buffering effects. The experimental and quasi-experimental studies investigate whether explicitly and/or implicitly observing animals is associated with direct effects on people’s physiology or associated with moderating people’s stress responses. Researchers have evaluated people’s responses to three different exposures to animals: (1) people explicitly looking at or observing animals or pictures of animals, (2) people implicitly observing or being in the presence of animals, and (3) people touching or interacting with animals.
6.4.1 Effects of explicitly looking at or observing animals or pictures of animals
Studies of the impact of looking at or observing animals document the direct impact of animals on people’s responses to scenes and the people in them (Lockwood, 1983; Rossbach and Wilson, 1992) and examine the physiologic indicators of parasympathetic nervous system arousal while and/or immediately after watching animals (i.e. Eddy, 1996, 1995; Globisch et al., 1999; Katcher et al., 1983). Only one research group (Katcher et al., 1983) addressed the effect of explicitly looking at or observing animals on people’s responses to stressors.
Friendly domestic animals have been used effectively in the advertising and publicity industries to impute safety, believability, and trustworthiness to people who accompany them (Lockwood, 1983). Research supports the positive influence of looking at animals on some of people’s moods and perceptions. Young adults rated scenes and the people depicted in pictorial scenes were rated as significantly more friendly (Lockwood, 1983), less threatening (Lockwood, 1983), happier (Lockwood, 1983; Rossbach et al., 1992) and more relaxed (Rossbach et al., 1992). In contrast pictures of animals culturally associated with fear elicited negative feelings and physiological arousal (Globisch et al., 1999).
Physiological indicators of parasympathetic nervous system arousal also indicate that looking at or observing domestic animals is associated with relaxation. Blood pressures of normotensive and hypertensive adults decreased progressively while watching fish swim in an aquarium (Katcher et al., 1983). The duration of the decreases was greater when observing an aquarium with fish than when looking at an aquarium with plants and moving water but without fish and than when looking at a wall.
Looking at familiar non-domestic animals can lead to decreases in physiologic arousal. The BPs and HRs of a chimpanzee’s caretaker and research assistants who assisted with the chimpanzee (n = 9) tended to be lower while watching the chimps than during a relaxation period (Eddy, 1995). In a single case report, the BP and HR of a 26-year-old male snake owner was lower during a six-minute period of watching his pet than during the preceding six minutes when he sat alone and relaxed (Eddy, 1996).
The potential stress response-moderating effects of watching animals first were in a study of the physiological impact of watching fish swim in an aquarium (Katcher et al., 1983). Blood pressure increases in response to reading aloud were less pronounced after watching fish than after watching other stimuli.
The studies of people observing fish and chimpanzees indicate that observing animals from a safe position often encourages people to relax. The constant motion of the animals studied in this context characteristically attracts the observer’s attention. The evidence presented through the comparison of the fish in the aquarium with the fishless aquarium and the wall support the contention that this attraction might be a prerequisite for continued relaxation over a longer time span (Katcher, 1981). Katcher suggested the biophilia hypothesis as one reason for people’s extended attention to the fish swimming in the tank compared with other stimuli. The data obtained during observation of loud, rambunctious chimpanzees suggest that profound tranquility and serenity might not be prerequisites for the decreased parasympathetic nervous system arousal while watching animals (Friedmann et al., 2000).