Physiological Effects of the Use of a Companion Animal Dog as a Cue to Relaxation in Diagnosed Hypertensives. Sue Todd Schuelke, RN, MSN, Brenda Trask, RN, MSN, Connie Wallace, RN, MSN, Mara M. Baun, DNSc., Nancy Bergstrom, Ph.D., RN, Barbara McCabe, Ph.D., RN
Reprinted with Permission from The Latham Letter, Winter 1991/92, The Latham Foundation, Latham Plaza Building, Clement and Schiller Street, Alameda, CA 9450l.
The purpose of this study was to determine if petting a companion dog to whom an attachment bond had been formed could serve as a cue for relaxation in diagnosed hypertensives. "Cue" in this study is defined as a stimulus that serves as a subconscious guide to an organism's response. One non-traditional relaxation method is petting or stroking a companion animal. It differs from other more widely accepted methods, such as biofeedback and progressive relaxation exercises, in that the attachment bond formed with the pet elicits the physiological responses indicative of relaxation rather than the extensive positive-negative feedback training involved in other therapies.
The human and pet are significant attachment figures for one another (Rynearson, 1978). Animals who are selected as pets are generally chosen because of an innate ability to display and respond to attachment. As a source of contact comfort, a pet can reduce its owner's anxiety and decrease sympathetic nervous system arousal. Pets can also provide a feeling of safety and a relaxing center of attention.
Some empirical evidence exists for the relaxation response to companion animals and the actual or potential health benefits for their human companions. Friedmann, Katcher, Lynch, and Thomas (1980) documented that 58% of the survivors surveyed 1 year post- myocardial infarction/angina pectoris had one or more pets. Of the patients who did not own pets, 23% died whereas only 6% of the pet owners died within the first year. The authors speculated that the contact comfort provided by the pets may affect the owner's physiological responses. Statistically significant decreases in both systolic and diastolic blood pressures have been reported in normotensive subjects (Baun, Bergstrom, Langston, & Thomas, 1984; Grossberg & Alf, 1985; Katcher, 1981).
In the only previous reported study using a small sample of hypertensives (10 subjects), there were statistically significant decreases over time in systolic and diastolic pressures when petting the subject's companion dog and statistically significant increases in finger temperature, perhaps an even better indicator of relaxation than blood pressure (Baun, Thoma, Langston, & Bergstrom, 1987). These results, while promising, represented findings from only a small sample. The question was asked if similar results would be obtained from a larger sample. Thus, it was hypothesized that petting a companion dog could serve as a cue for short-term relaxation of hypertensive individuals resulting in lower blood pressure and heart rate and increased finger temperature.
Design
A within-subject design was used to study the effects of petting a companion or control dog on the dependent variables of systolic and diastolic blood pressure, heart rate, and finger temperature.
Sample
Unpaid volunteer subjects were recruited from a midwestern metropolitan area by contacts with kennel clubs, hypertensive clinics, veterinarians, and acquaintances. Criteria for inclusion in the study were
- diagnosis of hypertension by a physician,
- 19 years of age or older,
- ability to give informed consent,
- owner of a companion dog to whom s/he was attached,
- willingness to come to the Behavioral Research Laboratory, and
- ability to speak and understand the English language.
The criterion for hypertension was the World Health Organization guidelines of systolic pressure 160 and diastolic pressure 90 mmHg. Pet attachment was confirmed by the Pet Attitude Scale. This study was approved by the Institutional Review Board (IRB) and the Animal Review Committee (ARC) of the University of Nebraska Medical Center.
Pet Attitude Scale
The Pet Attitude Scale was utilized to measure favorableness toward pets (Templer, Salter, Baldwin, & Veleber, 1981). The Pet Attitude Scale is an 18-item Likert format scale that has been found to have a Cronbach's alpha of .93 (p .001) and test-retest reliability of .92 (p .001). Construct validity was established by principal-component factor analysis. Range of possible scores is 38 to 266 with higher scores indicating more favorable attitudes.
Kendall blood pressure monitor
A Kendall blood pressure monitor model 8200 was used to measure systolic and diastolic blood pressure in millimeters of mercury (mmHg) and heart rate in beats per minute (bpm). The blood pressure monitor was fully automatic and noninvasive with reading displayed on digital readout windows. The blood pressure monitor was calibrated using a mercury manometer immediately prior to the study. The monitor contains a microcomputer that automatically calibrates a pressure transducer zero reference before measurement. Drift over time was negligible. Specifications indicated the heart rate range of the monitor was 30 to 300 bpm.
YSI telethermometer
Peripheral skin temperature was measured by the YSI telethermometer. The telethermometer was calibrated by the Biomedical Instrumentation Department immediately prior to the study by comparing the thermistor probe to a National Bureau of Standards' (NBS) calibrated Fahrenheit mercury thermometer.
The independent variable in this study was attachment to a companion dog. Since it had been found in an earlier study (Baun et al., 1984) that petting one's own dog to whom one was attached produced favorable effects on blood pressure, each subject's companion dog was used in this study. Because the active movement of petting could, of itself, increase blood pressure and heart rate in order to meet the metabolic demands of the muscles involved, the control protocol involved petting an unknown dog. In this way, the experimental and control protocols were identical in every way except for the element of attachment during the experimental treatment. Thus, it could be determined if petting one's dog could be therapeutic for hypertensives, or if some other factor was involved.
The subjects came with their dogs to the Behavioral Research Laboratory reception area. Prior to the study, informed consent was signed. In order to make each study session for each subject as similar as possible, the preliminary steps of gathering background information, presentation of instructions, and obtaining baseline and experimental conditions were presented in the same order and by the same investigator. Approximately 15-minutes were allowed for subjects to draw for random treatment order (own pet versus control) and complete the Pet Attitude Scale and a demographic data form so that blood pressure and heart rate had a chance to stabilize at a baseline level prior to the study. The subject was seated on a low, upholstered bench with a back so s/he was upright with feet flat on the floor. The monitoring equipment was attached to the subject's non-dominant arm which was positioned on a pillow on a low table.
The Behavioral Laboratory was equipped with a one-way mirror of which the subject was made aware and which facilitated observation of the subject's interaction with the dog. The blood pressure monitoring device and the finger temperature were fitted with long electrical cords and were located behind the one-way mirror to eliminate noise and distraction during the study. A laboratory setting was selected to control for extraneous variables which were previously found in the subject's home (e.g., door slamming, phone ringing, children playing) (Oetting, Baun, Bergstrom, & Langston, 1985).
The first random-ordered treatment began after obtaining 5-minutes of a baseline recording of blood pressure, heart rate, and peripheral skin temperature at 1-minute intervals. A period of 3 minutes of sitting quietly between treatments was allowed. The alternate animal was then brought into the room, positioned next to the subject, and the 12-minute measurement session was repeated. The control dog was a beagle who had been used in previous studies and was trained to sit or lie quietly next to each subject. Blood pressure, heart rate, and peripheral skin temperature were measured every 3 minutes during each treatment. Both subjects and dogs were under direct visual observation of the investigators at all times.
Data were analyzed using descriptive statistics, repeated measures ANOVA, and t-tests. Univariate repeated measures ANOVA was used on subjects' mean difference scores. The mean difference scores were calculated in order to eliminate individual differences between subjects and were determined by subtracting the subject's baseline value from each observed value. Baseline was defined as the mean of the five pretreatment recorded values. A significance level of p .05 was chosen for this study.
The sample consisted of 31 volunteer adult subjects; seventeen subjects were female and 14 were male. The ages ranged from 21 to 85 years (M = 50.01, SD = 13.2). Subjects had been diagnosed as hypertensive for a mean of 8 years (SD = 6.7) and had established a companion bond with their dogs as evidenced by a possible score of 257 on the Pet Attitude Scale (PAS) (M = 197.9, SD = 26.2). Subjects were being treated for hypertension as follows: (a) eight were being treated with diet; (b) one used exercise alone; (c) one used diet, exercise, and behavior modification; (d) three were treated with diet and dialysis; and (e) all but two were being treated with medication. Seventy-one percent of the subjects were treated with diuretics and/or adrenergic beta blockers. Nine of the subjects smoked.
There were no statistically or clinically significant differences for heart rate. The time effect was statistically significant for mean differences in systolic (F = 2.67, p = .049) and diastolic (F = 633.57, p = .0001) pressures. Protocol effects, i.e., companion dog versus control dog, however, were not statistically significantly different either for systolic or diastolic pressures. There was a statistically significant time effect for peripheral skin temperature (F = 4.29, p = .02). In addition, there was a statistically significant protocol effect (F = 3.78, p = .03) with a larger mean increase seen in the subjects during the experimental protocol.
Relaxation occurred in this study during petting subjects' companion animals as evidenced by increased peripheral skin temperature but not by decreased blood pressure as demonstrated in earlier studies. Since finger temperature has been used as an indicator of relaxation in a number of previous studies, it may be a more sensitive indicator of relaxation during petting of a companion dog than is blood pressure which can be affected by altered physiological pressures in hypertensives, i.e., renal failure, elevated renin-angiotensin levels, atherosclerosis, and the use of beta blockers.
A number of well-known relaxation techniques have been used to either decrease blood pressure or increase finger temperature or both. These techniques accomplish their effect by training the individual to exert control over the autonomic nervous system, a major modulator of blood pressure. Relaxation techniques, such as autogenic relaxation, biofeedback, progressive muscle relaxation, medication, self-hypnosis, and visualization, have been used to treat migraines, tension headaches, insomnia, hypertension, and Reynaud's disease (McEntee& Petticord, 1987; Rice, Caldwell, Butler,& Robinson, 1986). Using biofeedback with individuals exhibiting Reynaud's syndrome, a mean decrease of 50% in symptom activity has been associated with increases in hand temperature of 3 to 4 C (Carney, 1983). Muscle relaxation training has been reported to increase digital finger temperature by 4 F (Pender, 1984).
Thus, the mean increase in digital finger temperature of 2 F in this study during petting a companion dog is somewhat lower than those increases reported with other techniques but is still indicative of the effectiveness of dog petting. The increase of 2 F is similar to the increase of 26 F found in an earlier study of hypertensives (Baun et al., 1987).
The findings for blood pressure, on the other hand, were not entirely similar to those of earlier studies. In this study, there were no protocol differences between petting companion dogs with whom the subjects had formed a companion bond and petting an unknown dog as were found in an earlier study of normotensives (Baun et al, 1984). In the earlier study of hypertensives (Baun et al., 1987) and in two studies of normotensives (Grossberg& Alf, 1985; Katcher, 1981), decreases of 7.8, 4.55, and 7.2 mmHg were found for systolic and 4.3, 3.3, and 8.1 mmHg for diastolic blood pressures. Protocol differences in these studies may be responsible to some extent for the lack of congruence of findings with the present study. A more likely explanation is, however, that blood pressure is not a good indicator of relaxation in all groups of subjects. In the case of an altered physiological state such as hypertension, another variable like finger temperature may be a more reliable indicator. Since other studies of relaxation techniques by hypertensives (Pender, 1984, 1985) also have reported significant decreases in blood pressure, it may be that the one-time measurement session in a new and somewhat artificial setting was not adequate to detect differences. It should be noted that these other studies reported findings after a number of weeks of training in the technique used.
In addition, it was observed that blood pressure decreased in this study until the 6-minute observation point and then began to increase. Similarly, finger temperature increased until the 9-minute observation and then began to decrease. It may be that the study session was too long for the subjects to sustain the initial relaxation effect of dog petting. These same time-limited effects were observed in the earlier study of hypertensives where blood pressure decreased for 14-minutes and then began to increase while finger temperature increased for 12-minutes and then began to decrease (Baun et al., 1987).
Recommendations for further research include developing training sessions using the companion animal as a cue to relaxation and measuring the physiological effects over a period of weeks. Such extended training might result in a more sustained blood pressure decrease than the one-time, untrained sessions used in this study.
In a time when preventive and nonmedical treatment is receiving substantial attention, a method as simple as petting one's dog has obvious appeal. It is not suggested that petting one's dog replaces other proven treatments for hypertension, but that petting a dog to whom one is attached also can promote relaxation with minimal side effects and maximum enjoyment.
- Baun, M., Bergstrom, N., Langston, N.,& Thoma, L. (1984). Physiological effects of human/companion animal bonding. Nursing Research, 33, 126-130.
- Baun, M.M., Thoma, L., Langston, N.,& Bergstrom, N. (1987). Effects of petting a companion dog on selected physiological indices of relaxation in the known hypertensive. People, Animals, and the Environment: Celebrating New Partnerships. Delta Society, Renton, WA.
- Carney, R.M. (1983). Clinical applications of relaxation training. Hospital Practice, 18, (7), 83-94.
- Friedmann, E., Katcher, A., Lynch, J.,& Thomas, S.. (1990) Animal companions and one-year survival of patients after discharge from a coronary care unit. Public Health Reports, 15, 307-312.
- Grossberg, J.,& Alf, E. (1985) Interaction with pet dogs. Journal of the Delta Society, 2, 20-27.
- Katcher, A. (1981) Interactions between people and their pets: Form and function. In B. Fogle (Ed.), Interrelations between people and pets (pp. 41-66). Springfield, IL.
- Charles C. Thomas.
- McEntee, M.A.,& Petticord, K (1987). Coping with hypertension. Nursing Clinics of North America, 22, (3), 583-592.
- Oetting, K.S., Baun, M.M., Bergstrom, N.,& Langston, N.F. (1985). Petting a companion dog and autogenic relaxation effects on systolic and diastolic blood pressure, heart rate, and peripheral skin temperature. Journal of the Delta Society, 2, (1), 72.
- Pender, N.J. (1984) Physiological responses of clients with essential hypertension to progressive muscle relaxation training. Research in Nursing and Health, 7, 197-203.
- Pender, N.J. (1985) Effects of progressive muscle relaxation training on anxiety and health locus of control among hypertensive adults. Research in Nursing and Health, 7, 67-72.
- Rice, V.H., Caldwell, M., Butler, S.,& Robinson, J. (1986). Relaxation training and response to cardiac catheterization: A pilot study. Nursing Research, 35, (1), 39-43.
- Rynearson, E. (1978). Humans and pet attachment. Psychiatry, 133, 550-555.
- Templer, D.I., Salter, C.A., Baldwin, R.,& Veleber, D.M. (1981). The construction of a pet attitude scale. The Psychological Record, 31, 343-348.
