The effect of the presence of a traditional and non-traditional companion animal on blood pressure and heart rate reductions after termination of mild cognitive stressors was tested on 50 adult normotensive males and females. Blood pressures and heart rate were monitored continuously and measures recorded every minute, during task and rest periods in three conditions. Cognitive tasks employed to induce arousal included mental arithmetic, coding and a cancellation task. Each task was used within one of three conditions: pet absent, pet present and visual interaction allowed, pet present and tactual interaction allowed. To control for possible effects of task difficulty and condition sequence, each task was employed in each of the three conditions an equal number of times and each condition was presented as the first, second and third in the sequence an equal number of times. Data showed an increase in cardiovascular responses during cognitive tasks in all conditions, particularly when the pet was present. Repeated measures ANOVA indicate that reductions in blood pressure and heart rate measures during the rest period following termination of the cognitive stressor were significantly greater in the condition where a pet was present but no tactual interaction was allowed. Patting the animal resulted in a significant reduction in heart rate but blood pressure was not significantly different from when the pet was absent. Results indicate that the mere presence of an unknown friendly pet, traditional or non-traditional, can aid the reduction of blood pressure and heart rate following termination of cognitive stressors and this is independent of subjective reports, gender and general attitudes (positive versus neutral) towards pets.

KEY WORDS: Companion-animal, cardiovascular, cognitive, stressor, interaction.

A possible relationship between the animal-human bond and physiological well-being was first indicated in a study by Friedmann, Katcher, Lynch and Thomas (1980). They found that, one year following hospital discharge, coronary patients with companion animals had significantly lower mortality rates than non-pet owners. This was independent of gender, socio-economic status, social support and exercise. This result triggered research in the past decade in various directions. Pet owners and non owners of all ages were compared on a wide variety of psychological and physiological well-being variables. One of the very few Australian studies has confirmed the association between companion animal ownership and lower coronary risk (Anderson, Reid and Jenning, 1992). Generally, studies have shown some physical and psychological health benefits to pet ownership, particularly when accompanied with a strong positive attachment. For example, positive pet ownership is associated with happiness (Garrity, Stallones, Marx and Johnson, 1989) and better physical health in elderly women (Ory and Goldberg, 1983; Siegel, 1993), lower levels of psychiatric disturbance (Straede and Gates, 1993) and positive psychological effects on hospital staff (Brickel, 1980).

A number of experimental studies have investigated the stress reducing effect of the presence of and interaction with pets in terms of reduction of blood pressures and heart rate. Here, however, results have been inconsistent. Some physiological research indicates that the mere presence of a companion animal can significantly reduce blood pressure and this effect continues following removal of the animal (e.g., Riddick, 1985; Katcher, 1981), while others indicate that lowered blood pressure and heart rate occur only during tactual interaction with the animal (e.g., Vormbrock and Grossberg, 1988).

Jenkins (1986), comparing heart rate and blood pressures of 20 bonded dog owners both while reading aloud and petting their dogs, found that blood pressure but not heart rate was significantly lower during petting time. The subjects were not to speak during petting time, as this can increase blood pressure (Lynch and Thomas, 1980) but this also makes it difficult to disentangle the effects of the pet presence from the effects of the reading aloud condition.

Wilson (1987) reported that the presence of an unknown dog results in lower blood pressures and heart rate than when reading aloud, but not as low as quiet reading. While interacting with an unknown friendly dog was less stressful than reading aloud, it was more stressful than reading silently. Again the results were confounded by allowing subjects to talk to the dog. However, when Katcher's (1981) subjects rested, read aloud or actively petted and talked to their pets, their blood pressure was lower when interacting with their pets than when resting quietly.

Grossberg, Alf and Vormach (1987) used cognitive tasks to induce arousal in 32 normotensive male dog owning students, half of whom were tested with their pets present. Cognitive tasks for both groups increased blood pressures and heart rate but there was no difference between the pet present and pet absent groups.

While Friedmann, Katcher, Thomas, Lynch and Messent (1983) found that petting one's own dog resulted in a decreased blood pressure similar to quiet reading (but only for latter trials), Baun, Bergstrom, Langston and Thoma (1984) found that petting an unknown dog did not have the same effect and concluded that the pet effect may be dependant on bonding. Eddy (1996) also suggested that bonding may be an important factor, when the presence of a non-traditional pet (a snake) resulted in reduced cardiovascular activity in its owner. No study to date, however, has determined the effect of a nontraditional unfamiliar pet.

While these studies show little consistent evidence of a pet effect, there are procedural problems. Firstly, repeated measure procedures, with each subject serving as his/her own control are more sensitive to differences in conditions, rather than between groups comparisons, especially with smaller samples. Further, it is possible that the task itself, designed to require the subject's whole attention and induce stress, is likely to be a distraction from the presence of the animal. A more useful measure would be to measure latencies of physiological responses to return to baseline following termination of the stressor, or alternatively, to measure the degree of reduction of physiological measures during a rest period following termination of the stressor, in the presence or absence of a pet. In addition, the potential confounding of speech on measures needs to be controlled and the effects of visual versus tactual interaction disentangled. The present study attempts to address these methodological problems.

The aim was to determine the effect of the presence of an unknown friendly pet on physiological measures of normotensive adults following the termination of a cognitive stressor, using a repeated measures design. Because of the inconsistent past results, it is difficult to make predictions for the present study. Unlike most of the past research, the subjects in the present study include both male and female adults, pets owners and non owners and those with positive and neutral attitudes towards pets. In addition, traditional and non-traditional unfamiliar pets were employed, to determine any difference in the effect on cardiovascular responses.

METHOD

Subjects

A total of 50 normotensive adults, including 14 males and 36 females ranging in age from 26 to 50 years, served as subjects. This convenience sample included students, staff members and partners from a regional university. Participation was voluntary. Subjects were required to be medically fit and willing to undergo mild cognitive stressors. They were naive as to the aim of the experiment, and were merely told that it was to determine the effect of cognitive tasks on physiological changes. They were also told that participation could earn them up to $10, depending on how quick and accurate they were at completing the three tasks. This was to further motivate the subjects to perform and thus increase arousal, but in actuality, all participants received the $10 following completion of all data collection and were debriefed as to the true aim of the study. In addition, all received a brief report of findings upon completion of data analysis.

Materials

These included recording equipment and materials, two desks, three chairs, three cognitive tasks and three friendly companion-animals.

Equipment to measure physiological responses included a Critikon Dinamap Vital Signs Monitor, 845 x t, an automatically inflating device that uses the oscillometric method to measure simultaneously mean arterial pressure (MAP), systolic pressure (SP), diastolic pressure (DP) and heart rate (HR). The blood pressure measures are displayed numerically as a figure comparable to mmHg and heart rate is presented in beats per minute. The experimenter used recording sheets to record blood pressure and heart rate for each minute for each condition during the task and rest period.

Three different paper and pencil tasks of approximately five minutes duration included mental arithmetic problems, coding and a cancellation task. Each was presented on paper and clear instructions were given prior to testing. The mental arithmetic task included a combination of addition, multiplication and subtraction problems. The subject was required to write the answer next to the problem.

The coding task was similar to the Coding subtest in the Weschler Adult Intelligence Scale (WAIS) (1985 revision). The page consisted of boxes with numbers from 0 to 9 in the top half and nothing in the bottom half. At the top of the page, the boxes with numbers 0 to 9 in the top half had a specific symbol in the bottom half. Each number had its own symbol. The subject was required to write the appropriate symbol for each number, in the box below the number without skipping any.

The cancellation task involved a page of randomly generated numbers (3's, 6's, 8', and 9's), 40 rows with 77 digits per row. Subjects were required to circle the digit "6" whenever it came immediately following the digit "9". There was at least one instance in each line where this sequence occurred and several instances where it did not.

The three different tasks were employed in an attempt to control for a practice effect. Several colleagues of the researcher, who were not subjects, attempted the three tasks while their blood pressures and heart rates were monitored, and results confirmed that the tasks were equivalent in eliciting increased cardiovascular responses.

If subjects completed one task within the allotted 3 minutes, they were provided with a second page of the same task to complete. The order of the tasks within the experimental sessions was counterbalanced. That is, each of the three tasks was paired with each of the three conditions an equal number of times. The sequence of the conditions was also counterbalanced. Each of the three conditions was used as the first, second and third in the sequence of conditions within any experimental session a close to equal number of times. These condition sequences included the following: ABC; BCA; CAB; BAC; CBA; ACB.

Chairs were required for the experimenter, subject and research assistant. The three different pets included two small dogs and a kid goat. Only one pet was used at any one time. Each pet was used an approximately equal number of times. These were pets of the researcher and her colleagues and were recruited on the basis of friendly nature, small size and availability. Each pet was controlled by the research assistant. Each pet was introduced to coincide with the six condition sequences and the different task/condition sequences within experimental sessions. The first dog, a male Maltese cross named "Harley" was used for the first six experimental sessions, the kid goat named "Florence" for the next six sessions and the second dog, a female toy poodle named "Tammy", for the third six sessions. This sequence was repeated until all 50 subjects were tested. Each experimental session took up to 60 minutes, usually only three subjects were tested each day, and total testing was conducted over four weeks.

Procedure

Subjects were told that there would be three tasks to complete, with a rest period following each. They were instructed that each task would be timed and they were to complete them as quickly and accurately as possible without making errors. The amount of money they could earn, up to $10, would be determined by how quickly and accurately they completed each task.

The blood pressures and heart rate were monitored continuously in the left arm (or right, if the subject was left-handed). The arm with the blood pressure cuff was rested on the desk in a comfortable position and subjects were told to keep the arm as still as possible and not to speak during the entire session as it would interfere with the blood pressure and heart rate measures.

The researcher stated "I will be checking for changes in blood pressure and these measures will be taken automatically each minute. I will be recording these changes during the session. I will tell you when to begin the tasks and when to stop and rest. My assistant will be sitting at the end of the desk to observe and help with monitoring. Because she has her pet with her today, she may need to bring her pet into the room at some point. If so, do not let that bother you". Prior to entering the room for the experimental session the subject met the assistant researcher and the pet and it was determined that the subject was not aversive to the possibility of the pet's presence during part of the session.

Prior to beginning the first task, baseline measures were taken for three to five minutes. Sixteen different sequences of task and condition order were used to prevent confounding from the sequence of task and condition and possible differences in task difficulty. The researcher sat at a desk alongside the subject, monitoring the blood pressure and heart rate. These measures could not be seen by the subject.

The research assistant was present during all conditions. During the "pet absent" condition (Condition A) she read silently, and during both "pet present" conditions (Conditions B and C) she sat with the pet within interacting distance of the subject. This was to control for any effect of the presence of the research assistant confounding with the effect of the presence of the pet. The research assistant sat at the end of the desk, diagonally positioned from the subject.

During the rest period in Condition B (pet present), the research assistant held the pet in view of the subject, close enough for ease of visual and tactual interaction between pet and subject. The researcher said "You may wish to look at the pet while you are resting, but please don't pat her." During the rest period in Condition C the researcher permitted interaction with the pet if the subject wished, during the rest period, and stated "You may wish to pat the pet during this rest period".

All conditions were keep constant, with the only differences being the task, and the presence or absence of the pet. Following baseline measures, the research assistant quietly left the room and returned within a minute with the pet if the following condition required the pet to be present, or returned with a book if the following condition required the pet to be absent. At the end of the last minute of each rest period, she left the room and returned either with or without the pet, depending on the condition which was to follow.

Subjects were requested to refrain from speaking throughout the session to control for the possible effect of speech on physiological measures. For this reason also, subjects were requested to remain seated and refrain from making large arm movements. Arm movements were necessary, of course, during tactual interaction with the pet. All subjects complied with the instructions.

Measures which were analysed included those recorded three minutes after the beginning of each task and three minutes after completion of the task, that is, three minutes into the five minute rest period. These were chosen to allow time both for arousal to occur during the task and for sufficient reduction of cardiovascular responses during the rest period.

Following the three conditions, subjects were asked whether they felt more relaxed during the rest period when the pet was present or absent and whether or not they were pet owners. Attitudes towards pets were also gauged as positive, neutral or negative. Subjects were requested not to discuss the experiment with others until they received their payment in the mail, following all data collection.

RESULTS

Of the total sample of 50, 26 males and 10 females were pet owners, and 10 females and four males were not. Pet ownership was a good but not perfect indicator of positive pet attitudes. That is, a minority of owners indicated neutral attitudes towards their pets, while several non pet owners considered themselves to be pet lovers.

Table 1 shows the means and standard deviations for physiological measures taken two minutes prior to the task (referred to as baseline) three minutes into the task and three minutes into the rest period following the task, for each of the three conditions: pet absent (Condition A), pet present (Condition B) and pet present with tactual interaction allowed (Condition C). Measures during the task were generally slightly lower during Condition A (pet absent) than in the conditions where the pet was present. Whilst in all conditions, the measures decreased during the rest period following the task, this reduction was generally greater for conditions where the pet was present.

Repeated measures ANOVA for physiological measures using difference scores (during cognitive task/during rest period following task) showed a significant difference between conditions for systolic blood pressure (SPB), F(92,2) = 26.46, p < 0.000, diastolic blood pressure (DBP), F(98,2) = 17.14, p < 0.000, mean arterial pressure (MAP), F(98,2) = 11.44, p < 0.000 and heart rate (HR), F(98,2) = 7.48, p < 0.001. Table 2 shows results of planned comparisons of difference scores for measures between the three conditions. Results of two-tailed paired t-tests show that reductions in all measures were significantly greater in the two conditions where the pet was present (Conditions B and C) than when it was absent (Condition A). Comparison of measures in the two conditions where the pet was present showed a significant reduction on two of the measures, DBP and MAP where the pet was present but no tactual interaction was allowed (Condition B).

Because "during task" measures in Conditions B and C were higher than in Condition A, and this may have resulted in a spuriously greater difference in the task/rest measures for those "pet present" conditions, the data were re-analysed, using the lower "during task" measures from Condition A. Repeated measures ANOVA for these measures, using the difference scores, showed a significant difference between conditions for SBP, F(98,2) = 3.82, p = 0.025, DBP, F(98,2) = 4.57, p = 0.013, and HR, F(98,2) = 5.77, p = 0.004. MAP showed only a trend towards significance, F(98,2) = 5.77, p = 0.086.

Table 3 shows results of planned comparisons of difference scores as for Table 2, using these re-analysed data. Results show that blood pressures and heart rate were significantly lower during the rest period in Condition B (pet present), and MAP was close to significance. However, a comparison of Conditions A and C show only heart rate as significantly lower in Condition C where tactual interaction was permitted. A comparison of Conditions (B and C) show significantly greater diastolic blood pressure and mean arterial pressure during Condition C where tactual interaction with the pet was allowed.

In Condition C, when subjects could pat the pet if wished, 78% chose to pat the animal and 22% did not. There were no significant differences in physiological measures between gender, attitudes towards pets, and pet and non pet owners. However, those with positive attitudes, who chose to pet the animal when permitted (in Condition C) showed a trend towards a significantly greater decrease in heart rate than those with neutral attitudes (48,1,F = 3.8, p = 0.057).

In response to the researcher's question "Did you feel more relaxed during the rest period with the pet present or absent?" 50% said they felt more relaxed when the pet was present, 46% said they were more relaxed when the pet was absent, and 4% were "unsure". However, these subjective reports are not consistent with the physiological changes. That is, regardless of whether subjects considered themselves to be more relaxed in the presence of the pet during rest period, physiological arousal was significantly lower than when the pet was absent.

DISCUSSION

Results indicate that cardiovascular measures reduced significantly more during conditions where a pet was present and visual interaction only was allowed. This result was independent of subjects' gender, attitudes towards pets and subjective reports of feeling more relaxed in the absence of the pet. However, when tactual interaction with the animal was allowed, heart rate only was significantly reduced at the 3 minute recording time. Because all other factors were kept constant, it appears that tactual interaction, involving gross movement of the arm and hand resulted in the blood pressures remaining higher than when such interaction was not allowed. Perusal of individual data indicates this to be the case. Those subjects who chose not to pat the animal had similar blood pressures to those taken during their Condition B rest period.

To disentangle this potentially confounding factor of movement in future research, one would need to allow the same degree of movement in all conditions, including pet absent and pet present with visual interaction only. For example, the subjects could be required to stroke some inanimate object during these conditions.

The reduction in cardiovascular responses during the presence of a pet in the present experiment is consistent with results from several studies (e.g., Riddick, 1985; Katcher, 1981). However, the present results are contrary to those reported by Vormbrock and Grossberg (1988) who found that cardiovascular responses lowered only during tactual interaction with the animal. A measure of blood pressures after a period of tactual interaction longer than five minutes in the present study may have provided evidence of a reduction in those subjects who elected to pet the animal. The fact that heart rates had decreased significantly by the three minute period indicates that blood pressures would be likely to follow suit. Besides, they had already decreased in those subjects who chose not to pat the animal.

While it was not anticipated that the presence of the pet during an involving task would decrease blood pressures and heart rate, the finding of greater cardiovascular responses during the task in both conditions when the pet was present, than when it was absent, was unexpected. Some past studies found that the presence of pets reduced blood pressures and heart rate during cognitive stressors (e.g., Allen, Blascovich, Tomaka and Kelsey, 1991) and others (e.g., Alf and Vormbrock, 1987) found no difference. None, however, reported that the presence of pets actually increased such cardiovascular responses. In the present study it is possible that the presence of the pet during the task added to the arousal because it was an extra, potentially distracting stimulus in a situation where the subjects were already under pressure to perform. The subjects in the present study, motivated by monetary reward for good performance were possibly more highly aroused than those in previous studies. It is questionable, however, as to whether this effect could be generalised to a natural setting. During the rest period, however, the distracting stimulus of the animal was possibly more welcome and aided in the reduction in cardiovascular responses. In comparison, during the rest period in Condition A, when the pet was absent, subjects may have been more likely to ruminate on the preceding task, resulting in a relatively greater latencies in returning measures to baseline. In future laboratory research, an improvement in the procedure would be to bring in the animal following task completion, at the beginning of the rest period.

In Condition C, where subjects were permitted tactual interaction with the pet, almost a quarter of the subjects chose not to. Some of those who did pet the animal, however, had neutral attitudes towards pets. Those with positive attitudes who chose to pat the animal showed a trend towards a significantly greater decrease in heart rate than those with neutral attitudes. While the present results indicate that one does not need tactual interaction nor even to have a positive attitude towards pets in general, to gain the benefit of the pet's presence, these benefits may be greater if one's attitude is positive. Furthermore, contrary to Baun et al.'s (1984) suggestion, such benefits are not dependent upon bonding with a particular pet. Results were similar, regardless of whether the pet was traditional (a dog) or non-traditional (a kid goat). Perhaps the important factor here is the non-threatening nature of the pet. It is unlikely that the presence of an unfamiliar non-traditional pet such as a snake would result in the lower cardiovascular activity experienced by the highly attached owner of the snake, reported by Eddy (1996). Probably unfamiliar nontraditional pets with a degree of "cuteness" and a history of harmlessness are more likely to cause responses similar to traditional pets.

Much care was taken in the procedure to control for potentially confounding effects of task order, condition sequence and particular pet, by counter balancing both for task order and condition sequence as much as possible, by retaining the same pet within an experimental session and by ensuring that introduction and employment of each pet coincided with the range of condition and task sequences.

It is interesting that the subjects' perception of their state of arousal during rest periods was not necessarily reflected in their physiological measures. While almost half of the subjects reported feeling less relaxed in the presence of the pet during the rest period, the physiological measures belied this subjective report, indicating the subjective reports and objective measures may not always be consistent. Those who reported that they were less relaxed, actually showed lower arousal when the pet was present. Eddy (1996) also refers to incongruity of verbal reports and indices of stress. Why some subjects should feel less relaxed in the presence of the pet is uncertain, but it is possible that the degree of stimulation provided by the pet in this instance, during the rest period, might have been interpreted by the subject as resulting in a less relaxed state than when consciously attempting to relax when the pet was absent.

Results of the present study indicate that the mere presence of an unknown friendly pet, traditional or non-traditional, can significantly reduce physiological arousal in normotensive adults, following termination of a cognitive stressor, regardless of gender, general pet attitude and subjective reports. Measures were confounded by gross motor movements during tactual interaction, resulting in significantly reduced heart rate but not blood pressure, by the 3 minute recording period for those who elected to interact with the pet. There are difficulties in directly comparing results from various studies due to differences in subject groups and procedures. It would therefore be useful to conduct a larger scale, controlled experiment to determine the effect of known and unknown pets on physiological measures during and following cognitive and physical stressors using hypertensive and normotensive males and females of a wide range of ages.

Legend for Chart:

A - Measure
B - Condition A Base
C - Condition A Task
D - Condition A Rest
E - Condition B Base
F - Condition B Task
G - Condition B Rest
H - Condition C Base
I - Condition C Task
J - Condition C Rest

A             B         C         D         E         F
                        G         H         I         J

SBP        112.72    115.94    113.04    112.64    121.42
(mmHg)     (10.3)    (12.19)   (10.50)   (10.4)    (12.60)

                     110.60    112.80    122.58    112.22
                     (11.15)    (9.9)    (13.84)   (12.18)

DBP         60.02     64.04     60.52     59.20     68.76
(mmHg)     (11.2)    (12.03)   (10.97)   (10.80)   (12.38)

                      57.94     59.82     68.22     60.76
                     (10.74)   (11.8)    (12.55)   (11.19)

MAP         82.40     85.92     82.22     81.84     90.04
(mmHg)     (10.66)   (11.66)   (10.51)    (9.62)   (11.88)

                      80.26     82.04     88.88     82.04
                      (9.57)   (10.12)   (11.91)    (9.82)

HR          72.82     79.56     72.68     72.04     81.58
(bpm)      (10.55)   (11.40)   (10.02)   (11.20)   (10.66)

                      70.98     71.44     81.16     70.84
                      (9.54)    (9.86)   (11.10)    (9.15)

Legend for Chart:

A - Condition
B - Measure
C - t value
D - p

A             B           C        D

A vs B       SBP        -6.46    0.000
             DBP        -5.43    0.000
             MAP        -4.30    0.000
             HR         -3.02    0.004

A vs C       SBP        -6.08    0.001
             DBP        -3.49    0.001
             MAP        -2.12    0.039
             HR         -3.62    0.001

B vs C       SBP         0.28    0.706
             DBP         2.66    0.010
             MAP         3.63    0.001
             HR          0.28    0.783

Conditions: A - pet absent; B - pet present; C - pet
present/tactual interaction.

Legend for Chart:

A - Condition
B - Measure
C - t value
D - p

A               B           C         D

A vs B         SBP        -2.66     0.011
               DBP        -2.26     0.028
               MAP        -1.91     0.062
               HR         -3.03     0.004

A vs C         SBP        -0.97     0.337
               DBP         0.27     0.790
               MAP        -0.17     0.865
               HR         -2.83     0.007

B vs C         SBP         1.74     0.087
               DBP         2.69     0.010
               MAP         2.22     0.031
               HR         -0.24     0.814

Conditions: A - pet absent; B - pet present; C - pet
present/tactual interaction.

References

Allen, K.M., Blascovich, J., Tomaka, J. and Kelsey, J.M. (1991). Presence of human friends and pet dogs as moderators of autonomic response to stress in women. Journal of Personality and Social Psychology, 61(4), 582-589.

Anderson, W.R., Reid, C.M. and Jennings, G.L. (1992). Pet ownership and risk factors for cardiovascular disease. The Medical Journal of Australia, 157, 298-301.

Baun, M.M., Bergstrom, N., Langston, N.F. and Thoma, L. (1984). Physiological effects of human/companion animal bonding. Nursing Research, 33(3), 126-129.

Brickel, C.M. (1980). A review of the roles of pet animals in psychotherapy and with the elderly. International Journal of Aging and Human Development, 12(2), 119-128.

Corson, S.A. and Corson, E.O. (Eds.). (1980). Ethology and non-verbal communication in mental health. New York: Pergamon.

Eddy, T.J. (1996). R.M. and Beaux: Reductions in cardiac activity in response to a pet snake. Journal of Nervous and Mental Disease, 184, 573-575.

Friedmann, E., Katcher, A.H., Lynch, J.J. and Thomas, S.A. (1980). Animal companions and one-year survival of patients after discharge from a coronary care unit. Public Health Reports, 95, 307-312.

Friedmann, E., Katcher, A.H., Thomas, S.A., Lynch, J.J. and Messent, P.R. (1983). Social interaction and blood pressure: Influence of companion animals. Journal of Nervous and Mental Disease, 171(8), 461-465.

Garrity, T.F., Stallones, L., Marx, B.M. and Johnson, T.P. (1989). Pet ownership and attachment as supportive factors in the health of the elderly. Anthrozoos, 3(1), 35-44.

Grossberg, J.M., Alf, E.F. and Vormbrock, J.K. (1987). Does pet dog presence reduce human cardiovascular responses to stress? Anthrozoos, 2(1), 38-44.

Jenkins, J.L. (1986). Physiological effects of petting a companion animal. Psychological Reports, 58(1), 21-22.

Katcher, A.J. (1981). Interactions between people and their pets; form and formation. In B. Fogle (Ed.) Interrelations between people and pets. Springfield, IL: Charles C. Thomas.

Lynch, J.J. and Thomas, S.A. (1980). Human speech and blood pressure. Journal of Nervous and Mental Disease, 168, 526-534.

Ory, M.G. and Goldberg, E.L. (1983). Pet possession and well-being in elderly women. Research on Aging, 5(3), 389-409.

Riddick, C.C. (1985). Health, aquariums and the non-institutionalized elderly. Special Issue: Pets and the family. Marriage and Family Review, 8(3-4), 163-173.

Siegel, J.M. (1993). Companion animals: In sickness and in health. Journal of Social Issues, 49(1), 157-167.

Straede, C.M. and Gates, G. (1993). Psychological health in a population of Australian cat owners. Anthrozoos, 6(1), 30-42.

Vormbrock, J.K. and Grossberg, J.M. (1988). Cardiovascular effects of human-pet interaction. Journal of Behavioral Medicine, 11(5), 509-517.

Wilson, C.C. (1987). Physiological responses of college students to a pet. Journal of Nervous and Mental Disease, 175(10), 606-612.

(Received 24 April, 1998; in final form 21 September, 1998)

~~~~~~~~

By Lesley R. Demello, School of Behavioural and Social Sciences and Humanities, University of Ballarat.