Peter Major
Источник: ELSEVIER©
Abstract
Many authors have considered heart rate as a way of assessing the sensitivity of animals to environmental challenges. Some environmental challenges may have no easily measured, overt, behavioural effects yet may still affect physiological measures, including heart rate. However, heart rate is primarily affected by metabolic demand (19). Therefore an animal’s physical activity, whether externally visible (e.g., locomotor behaviour) or not (e.g., muscle tension), affects heart rate; as the rate of energy expenditure increases, heart rate increases (25,29–31). Posture also affects heart rate (6,29); although this effect is not entirely due to changes in metabolic demand, for the purposes of this study I shall include posture as a component of physical activity.
The idea of somatic coupling [see (19,20)] hypothesised that all changes in heart rate are mediated via changes in metabolic demand, so that only psychological factors that affect physical activity, though this activity might not be externally visible, can be assessed by measuring heart rate. It was subsequently demonstrated (3) that heart rate is increased by anxiety, and is not solely a measure of physical activity. This, however, does not remove the primary dependence of heart rate upon metabolic demand. In several studies of both humans and other primates, heart rate has been measured to infer some aspect of the psychological state of an animal. Some of these studies compare heart rate across different environmental conditions [e.g., (4,5,7,8,24,27,28)], others across different groups of individuals [e.g.,(12,14,21,23,32,34)]. Although many studies acknowledge that there is a relationship between physical activity and heart rate, and although attempts have often been made to describe, measure, or control physical activity, this has, on occasions, been done without considering how heart rate varies across a range of physical activities [e.g., (23,24)]. Frequently, an implicit assumption has been made that whole body movements through space, which I refer to as locomotion, are of predominant importance in determining heart rate, whereas other forms of physical exercise, such as grooming, masticating, scratching, standing still, or hanging from the walls of a cage, none of which involve locomotion, do not require consideration.
Some studies of nonhuman primates have examined the relationship between heart rate and physical activity. The heart rate of pigtail macaque infants was found (22) not to correlate within individuals either with the duration of motion (where motion was desined as walking, running, or climbing) or with activity per unit time (where activity was de?ned as movement across imaginary subdivisions between equal-sized fractions of the enclosure). Variation in heart rate between individuals was found (8) not to correlate with the number of 15-s intervals in which three species of macaque locomoted and/or manipulated a chest harness. Heart rate was found (26) to correlate positively within individuals with a count of periods in which locomotion occurred in adult squirrel monkeys, where locomotion was de?ned as physically relocating the body by at least one body length.
However, the apparently contradictory nature of these findings, both with respect to each other and, in some cases, with respect to our expectation of the effect of physical activity upon heart rate, along with the failure of these studies to measure nonlocomotor movements, means that our understanding of the relationship between heart rate and physical activity of nonhuman primates is poor.
This in turn limits the usefulness of heart rate as a tool in investigations of psychobiological effects and has sometimes led to misinterpretation of heart rate data. Therefore, I attempted to assess the interval between successive heart beats, which I refer to as interbeat interval (IBI), over a range of different physical activities, including both nonlocomotor and locomotor, commonly shown by cage-living rhesus macaques. This allowed an explanation of those unexpected endings mentioned earlier; additionally, it necessitated the reevaluation of some previous inferences of psychobiological effects upon heart rate.
Method
I used 14 rhesus macaques: seven multiparous mothers, aged 7–16 years, and each mother’s youngest infant, aged 13 weeks. Four of these infants were male; three were female. Each mother infant pair was transferred from its home group into an environment affording visual, olfactory, and auditory isolation from the other animals in the colony. Pairs were housed in a cage, 1.2 *1.2 * 1.5 m, with cable mesh front, sides, ?oor, and ceiling and a solid back. The cage was ?tted with a solid shelf, food dispenser, water bottle, and radiotelemetry receiver. Animals were fed twice daily, at 0900 hours with pelleted monkey food, shelled peanuts, and sun?ower seeds and at 1400 hours with fresh fruit or vegetables. The room was maintained at a temperature of 18 ± 2.5°C; lighting of 350 lux at bench level was provided from 0700 to 1900 hours.
Each animal was fitted with a radiotelemetry transmitter, size 35 * 70 *10 mm, built at the Sub-department of Animal Behaviour in Madingley. Mothers were anaesthetised using ketamine (10 mg/kg intramuscular); the transmitter was implanted at a subcutaneous, dorsal site at least 12 days prior to observation. Infants were lightly anaesthetised; the transmitter was externally mounted in a ventral position and covered with a flexible PVC jacket 5 h prior to observation. The jacket allowed a complete range of infant locomotor activities while protecting the transmitter from grooming by either the infant or its mother; grooming was not increased by the presence of the jacket. Due to problems of tissue leakage into some implanted telemetry transmitters, data from two of the mothers are unavailable.
I collected data from mothers and infants by simultaneouslymonitoring animals for 74 ±5 (mean ± 1 SEM) min between 1425 and 1735 hours on each of two afternoons 3 days apart. Data from these two afternoons were combined, as no interaction between day and IBI was found. I recorded activity onto videotape from a neighbouring room using closed-circuit television; radiotelemetry signals were synchronously recorded onto the audio channels of the videotape.
The physical activity of animals during the observation periods was subdivided into six categories:
- Sitting still (SitS): Sitting with no readily observable body movements, excluding inhalation and exhalation.
- Sitting moving (SitM): Engaging in body movements (e.g.,scratching, grooming, masticating, or irregular movements of the head) while sitting.
- Standing (Stand): Standing, either with or without body movements, either on all four limbs or on hind limbs with some of the body weight supported by the forelimbs against the side of the cage.
- Hanging (Hang): Hanging either from the walls of the cage or, in infants, from the mother. Walking (Walk): Walking.
- Locomoting (Locom): Locomotion other than walking (e.g., running, leaping, or climbing up, across, or down the cage).
These six categories include all the activities observed in the monkeys. The categories were defined to allow enough suficiently prolonged periods of observation of each animal in each activity. A greater number of more narrowly defined categories was not used because longer observation periods would then have been necessitated; this would have increased the problem of state changes within observation periods.
The activity of each monkey was divided into periods of each category of activity. All periods of ≥11 s were used in subsequent analysis. IBI was sampled for one 5-s period within the ≥11-s period. Sampling began ≥5 s after the onset of the activity and ended $1 s before the onset of the following activity; whenever possible the period sampled was that ending 5 s before the onset of the following activity.
This sampling regime was adopted to reduce the effects upon heart rate of both how long an animal had been engaged in an activity, and how long in the preceding activity. To describe such effects in detail is beyond the scope of this study: it would require observation over long periods, in which extensive state changes might confound the subsequent analysis. Other studies, not of rhesus macaques, have found that cardiac acceleration occurs almost instantaneously at the onset of exercise (15,33). With prolonged locomotion, heart rate is either maintained at a steady level (2,10,15) or increases (11). Upon the cessation of exercise, heart rate remains elevated for 10–30 s (10,15,16,33).
In my procedure, sampling ceased 1–5 s before the onset of a subsequent activity. However, in other species, increases in heart rate have been observed before the onset of locomotion (9,10). If preactivity effects were demonstrated in the rhesus macaque, the results presented here might require reconsideration.
Within each 5-s period, the telemetry signal was recovered from the videotape using a frequency discrimination circuit. High-frequency artefacts in the electrocardiograph record were edited at the analogue level prior to determining the mean IBI within the 5-s period. The mean IBI for each category of activity was then determined. IBI is the reciprocal function of heart rate and was used in this study due to the high sampling error incurred when measuring heart rate within short time periods. Data were analysed using repeated-measures ANOVA. A post hoc analysis, again using repeated-measures ANOVA, was made to investigate whether sitting moving has a different effect upon IBI from sitting still. Not all animals showed all categories of activity; therefore the maximum sample size of twelve (move mothers and seven infants) is reduced for some comparisons.
Results
Physical activity affected IBI differently in mothers and in infants: there was an interaction between group and activity category (n = 7, df = 5, F = 5.269, p = 0.002); see Fig. 1. Therefore the data from mothers and infants were considered separately.
All move mothers showed four categories of activity: SitS, SitM,Stand, and Walk. IBI varied across these activities (n =5, df = 3, F = 22.618, p ≤ 0.0001); see Fig. 1. Body movements while sitting reduced IBI from that when sitting still (n = 5, df = 1, F =50.876, p = 0.002), by an average of 17%. Three mothers exhibited all six categories of activity; in these animals IBI still varied with activity (n = 3, df = 5, F = 13.137, p = 0.0004). Locomoting decreased IBI by an average of 45% from that when sitting still.
All seven infants showed four categories of activity: SitS, SitM, Stand, and Hang. IBI varied across these activities (n = 7, df = 3, F = 48.892, p ≤ 0.0001); see Fig. 1. Body movements while sitting reduced IBI from that when sitting still (n =7, df = 1, F = 32.978, p = 0.001), by an average of 18%. Fourinfants exhibited all six categories of activity; in these animals IBI still varied with activity (n = 4, df = 5, F = 29.629, p ≤0.0001). Locomoting decreased IBI by an average of 31% fromthat when sitting still.
Discussion
The major endings of this study are threefold. Firstly, theeffects of physical activity upon heart rate are large. Secondly, apparently similar activities of mothers and infants affect their heart rates differently: in mothers there is generally a greater change than in infants. This may result from effects due to the size or age of an animal, which may arise due to differences in metabolic demand, cardiac physiology, or behaviour. This ending implies that a description of the relationship between physical activity and heart rate in one group of animals may not be extrapolated precisely to a different group. Thirdly, different physical activities have different effects upon heart rate. The ending that, in caged primates, even commonly observed, nonlocomotor activity has a large effect upon heart rate requires that several published studies be reconsidered.
Studies of other animals have found that heart rate is sensitive to relatively subtle variation in physical activity. In humans, there is an approximately linear relationship between the intensity of work in watts and heart rate (29). Among ungulates, both free-ranging sheep (16) and white-tailed deer (13,18) show a series of increases in heart rate as they change activity from bedded to standing to walking to trotting and running. Walking downhill results in a lower heart rate than walking on the level or uphill (16). The effects on sheep heart rate of different activities are additive: the increase in heart rate associated with standing and ruminating is equal to the sum of the increases associated with each activity separately (1). In gulls, looking around or preening causes a ~20% increase in heart rate above that while resting, soaring ?ight a ~20% increase, and ?apping ?ight a ~100% increase (15).
However, many studies of primates have used relatively unsubtle categorisations of physical activity. Having now demonstrated that such categorisations are incomplete, the endings can be reevaluated. As described in the introduction, within-individual correlations of heart rate with either the duration of motion or an activity count were not detected in pigtail macaque infants (22). However, both of these measures are indicators solely of activity involving locomotion; they are insensitive to variation in other types of activity. Day-to-day variation in the time spent in nonlocomotor activities has now been shown to affect heart rate. Such variation can occur independently of variation in locomotion; therefore the expected increase in heart rate on days when locomotion was increased might not have been seen. The absence of a significant between-individual correlation of heart rate with the number of 15-s periods in which macaques locomoted and/or manipulated a chest harness (8) may now be explained in a similar way: some variations in physical activity that affected heart rate may not have been observed when using this particular measure of activity.
The large and unequal effects of physical activity upon heart rate are of consequence when researchers attempt to infer psychological effects upon heart rate. Even subtle changes in physical activity might explain changes in heart rate previously ascribed to psychological factors. This is not to suggest that all changes in heart rate are mediated via changes in physical activity; however, physical activity must be accounted for before inferring psychological effects. Three ways of designing studies to achieve this are by controlling against variation in physical activity, by determining the correlation between physical activity and heart rate, and by correcting for physical activity; each of these approaches allows different types of behaviour to be examined (17).
Some previous studies in this field either failed to consider physical activity or differentiated physical activity into too few or inappropriate categories; these should be reconsidered in the light of the present endings. For example, in 20- to 30-week-old pigtail macaques, following maternal separation in the presence of peers, the change in heart rate positively correlated with the change in peer contact (7). This may be due to psychobiological factors, such as the inability of infants showing low arousal to maintain close peer bonds, or the tendency of infants showing high arousal to seek peer contact. Equally, however, it may be due to those infants that spend more time in contact with peers being more active, possibly due to locomotion associated with play, and therefore having higher heart rates. To test these hypotheses, further data describing infants’ physical activity are required.
In conclusion, the present study showed that even activities that do not involve locomotion result in decreases in IBI. In the absence of detailed descriptions of an animal’s activity, little can be deduced about the relationship between an animal’s behaviour and its heart rate. Future studies of heart rate should take care to account for effects of physical activity, including nonlocomotor, before inferring psychological effects.
References
Baldock, N. M.; Sibly, R. M.; Penning, P. D. Behaviour and seasonal variation in heart rate in domestic sheep, Ovis aries. Anim. Behav.36:35–43; 1988.
Barnard, R. J.; Duncan, H. W.; Thorstensson, A. T. Heart rate responses of young and old rats to various levels of exercise. J. Appl. Physiol. 36:472–474; 1974.
Blix, A. S.; Stromme, S. B.; Ursin, H. Additional heart rate—An indicator of psychological activation. Aerosp. Med. 45:1219–1222; 1974.
Boccia, M. L.; Reite, M.; Laudenslager, L. On the physiology of grooming in a pigtail macaque. Physiol. Behav. 45:667–670; 1989.
Cacioppo, J. T. Social neuroscience: Autonomic, neuroendocrine, and immune responses to stress. Psychophysiology 31:113–128; 1994.
Cacioppo, J. T.; Uchino, B. N.; Berntson, G. G. Individual differences in the autonomic origins of heart rate reactivity: The psychometrics of respiratory sinus arrhythmia and preejection period. Psychophysiology 31:412–419; 1994.
Caine, N.; Reite, M. The effect of peer contact upon physiological response to maternal separation. Am. J. Primatol. 1:271–276; 1981.
Clarke, A. S.; Mason, W. A.; Mendoza, S. P. Heart rate patterns under stress in three species of macaques. Am. J. Primatol. 33:133–148; 1994.
Dowell, R. T. Cardiovascular physiology of exercise. In: Bove, A. A.; Lowenthal, D. T., eds. Exercise medicine: Physiological principles and clinical applications. Orlando: Academic Press; 1983:19–27.
Dressen, W.; Hendrichs, H. Social behaviour and heart rate in tammar wallabies (Macropodidae: Macropus eugenii). J. Zool. 227:299–317; 1992.
Ekelund, L. G.; Holmgren, A. Central hemodynamics during exercise. Circ. Res. 20:I-33–I-43; 1967.
Field, T. M. Early interactions between infants and their postpartum depressed mothers. Infant Behav. Dev. 7:517–522; 1984.
Jacobsen, N. K. Changes in heart rate with growth and activity of white-tailed deer fawns (Odocoileus virginianus). Comp. Biochem. Physiol. 62A:885–888; 1979.
Kagan, J. The concept of behavioral inhibition to the unfamiliar. In: Reznick, J. S., ed. Perspectives on behavioral inhibition. Chicago: The University of Chicago Press; 1989;1–23.
Kanwisher, J. W.; Williams, T. C.; Teal, J. M.; Lawson, K. O. Radiotelemetry of heart rates from free-ranging gulls. Auk 95:288–293; 1978.
MacArthur, R. A.; Johnston, R. H.; Geist, V. Factors in?uencing heart rate in fre e-ranging bighorn sheep: A physiological approach to the study of wildlife harassment. Can. J. Zool. 57:2010–2021; 1979.
Major, P. Heart period and other physiological correlates of the behaviour of mother and infant rhesus macaques. Cambridge: University of Cambridge; 1995 (Unpublished doctoral dissertation).
Moen, A. N. Seasonal changes in heart rates, activity, metabolism, and forage intake of white-tailed deer. J.Wild. Manage. 42:715–738; 1978.
Obrist, P. A. Cardiovascular psychophysiology. New York: Plenum Press; 1981.
Obrist, P. A.; Webb, R. A.; Sutterer, J. R.; Howard, J. L. The cardiacsomatic relationship: Some reformulations. Psychophysiology 6:569587; 1970.
Rasmussen, K. L. R.; Fellowes, J. R.; Suomi, S. J. Physiological correlates of emigration behavior and mortality in adolescent male rhesus monkeys on Cayo Santiago. Am. J. Primatol. 20:224–225; 1990.
Reite, M.; Short, R. Circadian rhythms in monkeys: Variability and behavioral correlations. Physiol. Behav. 27:663–671; 1981.
Reite, M.; Short, R. Behavior and physiology in young bonnet monkeys. Dev. Psychobiol. 19:567–579; 1986.
Reite, M.; Short, R.; Seiler, C.; Pauley, J. D. Attachment, loss and depression. J. Child Psychol. Psychiatry 22:141–169; 1981.
Rushmer, R. F.; Smith, O. A. Cardiac control. Physiol. Rev. 39:41–68; 1959.
Saltzman, W.; Mendoza, S. P.; Mason, W. A. Sociophysiology of relationships in squirrel monkeys. I. Formation of female dyads. Physiol. Behav. 50:271–280; 1991.
Shapiro, P. A.; Sloan, R. P.; Horn, E. M.; Myers, M. M.; Gorman, J. M. Effect of innervation on heart rate response to mental stress. Arch. Gen. Psychiatry 50:275–279; 1993.
Sloan, R. P.; Korten, J. B.; Myers, M. M. Components of heart rate reactivity during mental arithmetic with and without speaking. Physiol. Behav. 50:1039–1045; 1991.
Smith, J. J.; Kampine, J. S. Circulatory physiology—The essentials. Baltimore: Williams & Wilkins; 1990.