Hazard rates and causes of death in a captive group of crab-eating monkeys (Macaca fascicularis).код для вставкиСкачать
American Journal of Primatology 30:13%147 (1993) Hazard Rates and Causes of Death in a Captive Group of Crab-Eating Monkeys (Macaca fascicufaris) HANS U. LUDER Department of Oral Structural Biology, Dental Institute, Uniuersity of Zurich, Zurich, Switzerland In an attempt to examine possible associations between stages of agespecific mortality and various causes of death, vital records of 159 male and 192 female crab-eating monkeys (Mucaca fascicularis), housed as a single group, were analyzed. Survival and hazard rates associated with each of five distinct categories of causes of death were estimated for males and females, using the nonparametric kernel method. The obtained overall survival and hazard functions were similar to those reported previously for rhesus monkeys. Among two stages identified in age-specificmortality, the first stage, characterized by rapidly decreasing hazard rates up to about 1.5-2 years of age, was discriminated by the occurrence of deaths due to unfitness for postnatal life. The second stage lasted up to the age of 10-15 years and was largely characterized by a high incidence of violent deaths. The respective hazard rates in males and females attained peaks during the early reproductive period of life and markedly decreased thereafter. This pattern was interpreted to indicate that second stage mortality is unlikely related to senescence, but rather, seems to depend on extrinsic environmental factors. Thus, when considering overall hazard rates in Mucacu fusczculuris, the onset of senescence, as a result of the specific aspects of simian reproduction, may be hidden from view, and mortality due to aging may only be appreciable after 10 to 15 years of age. 0 1993 Wiley-Liss, Inc. Key words: Macaca fascicularis, survival, hazard rate, causes of death, type of mortality INTRODUCTION In search for an appropriate animal model for various aspects of human aging, several studies of survival and mortality in nonhuman primates, mostly captive rhesus monkeys, have been undertaken during the last decade [Smith, 1982; Rawlins et al., 1984; Dyke et al., 1986; Gage & Dyke, 1988; Tigges et al., 19881. In an attempt to develop a mortality standard for larger Old World monkeys, Gage and Dyke  fitted the Siler model to data sets obtained from eight, mostly captive populations of three monkey species. The Siler model is a three component com- Received for publication February 13, 1991; revision accepted January 9, 1993. Address reprint requests to Dr. H.U. Luder, Dental Institute University of Zurich, Department of Oral Structural Biology, Plattenstrasse 11, CH-8028 Zurich, Switzerland. 0 1993 Wiley-Liss, Inc. 140 I Luder peting hazard function allowing for a n immature and a senescent, in addition to a residual, component of mortality. The latter component is assumed to be due to “causes of death so severe that they kill regardless of age” and, therefore, to be constant throughout life. Hence, residual mortality should be most evident during a n age period when the immature and senescent components of mortality are low, i.e., during adolescence and early adulthood. However, when applying the Siler model, Gage and Dyke I19881 did not obtain significant values for the residual component except in one animal group. This finding suggests that a residual hazard is essentially missing from populations of captive monkeys. Consequently, a n early period characterized by deaths from causes associated with immaturity would be continuous with a later period distinguished by causes of death due to senescence. It was the purpose of this study to test the above suggestion that simian mortality lacks a distinct third component competing with the immature and senescent components. This was done by examining age-specific hazard rates and causes of death in a group of captive monkeys. METHODS Vital records of 159 male and 192 female monkeys (Macuca fuscicularis) were used for analysis. Animals were born between January 1,1961 and December 31, 1989 and belonged to a colony kept in the zoo of Base1 (Switzerland). Over the entire observation period, a maximum of 80-90 animals lived together a t any one time and were housed day and night as a single group. The animals did not receive any special daily care except food and water. Disease prophyllaxis, such as vaccination against specific diseases, or routine veterinary care were not done. Medical attention was confined to obvious illness and injuries. In such cases, animals were isolated, usually for 1-2, a t most for 5 days, and then returned to the colony. With few exceptions, these monkeys were reintegrated in the group and survived to breed. Euthanasia was resorted to when a disease or injury proved resistant to therapy or was considered by the veterinarian to be too severe to be cured. However, no animals, either newborn or very old, were killed so as to reduce the group size. Population size was maintained by culling and removing animals to other colonies. Vital records included the exact dates of birth, of removal from the colony, and of death, as well as of diseases and medical interventions. Euthanized animals were classified as having died; stillbirths were excluded from the analysis. Autopsy and microbiological tests were performed routinely on deceased animals except in a few cases where advanced autolysis or severe mutilation of the cadavers prevented a postmortem examination. Based on the autopsy reports, causes of death were assigned to the following categories: (1) “Immaturity” was assumed, when animals apparently died, because they were biologically unfit to cope with the demands of postnatal life. Thus, deaths resulting, for instance, from severe birth defects a s well a s from maternal neglect and subsequent starvation were summarized under this heading. (2) “Violence” involved fights as well as accidents. In cases where the violent event had not been observed directly, it was surmised based on the nature of the injury evinced. Thus, bites were always assumed to have been incurred during fights, while blunt trauma rather indicated a n accident. Clearly, these criteria did not allow the determination whether a n accident had occurred in the course of a fight. Hazard Rates a n d Causes of Death I 141 (3) “Infections” were indicated by positive results of microbiological tests and comprised infectious diseases as well as sepsis associated with severe injuries. (4) “Other causes” included those which could not be assigned to one of the above categories. (5) An “unknown cause” was assumed when postmortem examination did not allow a clear decision or when no autopsy could be performed. In cases where two of the above possibilities could be considered, such as combinations of severe injuries and infections, the cause of death was partitioned equally to both categories. Survival and mortality data of male and female animals were summarized in separate life tables [Miller, 19811. From these life table values, smooth survival and hazard functions were derived using the nonparametric kernel method and a commercially available collection of FORTRAN subroutines (CURVDAT from STATCOM, Neckargemund, Germany) running on a Micro PDP-11/73 (digital) computer [Gasser & Muller, 1979; Miiller & Wang, 19901. While survival functions were obtained directly from the life table data (i.e., as the zero derivative), hazard rates were estimated as the first derivative of the cumulative hazard values. Kernels of order 2 (for zero derivatives) and 3 (for first derivatives) and fixed bandwidths were used for smoothing. The latter were selected so as to represent a compromise between data-based optimal bandwidths for the curve segments with the highest and lowest density of data points. By applying this procedure separately to the respective life table data, distinct curves of hazard rates against age were obtained for each cause of death. RESULTS Most deaths of both male and female animals occurred within the first 3 months following birth (Tables I and 11). Thereafter, mortality rates dropped sharply. Although one male and one female attained ages of 23 and 28 years, respectively, the oldest deceased male was 12 and the oldest deceased female 26 years of age. Most animals removed from the colony and lost to follow-up ranged in age from 1 to 10 years. As a result, the number of living males and females older than about 15 and 20 years, respectively, was low. Survival rates of male and female animals up to the age of about 7 years were similar and characterized by a steep decline immediately after birth and a subsequent slow, though continuous, decrease (Fig. la ). Corresponding to a steep increase in hazard rate, the proportion of surviving males dropped markedly after the age of about 7-8 years. By contrast, the hazard rate for female monkeys remained low and started to increase only at the age of about 1 4 years. Most deaths within the first 2 years of life in both genders were caused by “immaturity,” “violence,” “infection,” and “unknown causes” (Fig. lb,c). Violent deaths a t this early age usually resulted from deadly accidents, mostly falls, and occurred somewhat more often in females than in males. Most of the infections were related to diseases, e.g., pneumonia or gastrointestinal diseases, rather than to severe injuries. During the period from about 2 to 12 years of age, “violence” in both genders prevailed over the other causes of death. The respective hazard rates attained a maximum around 8 years of age in females (Fig. lb) and two peaks around 5 and 10 years of age in males (Fig. lc). After these peaks, the risk of dying from violence dropped sharply. Violent events which occurred early in this age period were partly accidents and partly fights, while violent deaths at higher ages were related to injuries incurred during fights. In several instances, such injuries were associated with sepsis. Thus, unlike infections occurring in the neonatal and 142 I Luder TABLE I. Life Table for Male Animals Number of animals Alive a t beginning of interval Age interval (years) 0-0.25 0.25-0.5 0.5-0.75 0.75-1 1-2 2-3 3 -4 4-5 5-6 6-7 7-8 8 -9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18-19 19-20 20-21 21-22 22-23 23-24 24-25 ~ 159 139 137 136 131 118 99 78 68 54 48 31 24 17 13 8 7 7 7 6 3 2 2 1 1 1 1 0 Censored within interval Exposed to risk of dying Died within interval 0 1 0 3 9 18 18 10 11 6 16 6 6 3 4 0 0 0 1 3 1 0 1 0 0 0 1 159 138.9 137 134.7 126.2 109.1 91.5 72.3 62.6 51.5 40.1 28.1 21.2 14.4 10.9 8 7 7 6.7 5 2.7 2 1.6 1 1 1 0.3 20 1 1 2 4 1 3 0 3 0 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 Died from violence 2 0 0 0 1 1 0 0 2 0 0 0 0.5" 1 1 0 0 0 0 0 0 0 0 0 0 0 0 Died from infection 12 0 1 1 1 0 1 0 0 0 0 0 0.5" 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ~ "In these cases, no conclusive decision could be made as to one particular cause of death infant period, infections recorded between 2 and 12 years of age in almost half of the cases were related to severe injuries. DISCUSSION The present analysis was undertaken to determine whether particular causes of death are associated with specific periods of the life cycle and could thus be used to characterize the respective stage in age-specific mortality. The chosen categories of causes of death compare to those based on the International Classification of Diseases, which were used for analyzing human mortality [Steffen et al., 19881. In agreement with this classification, fights as well as accidents were considered violent events. Thus, the category ((violence"comprised merely instantaneous deaths due to trauma, as against deaths resulting from diseases or other natural causes. In addition to the classification used for causes of death in humans, a category termed "immaturity" was introduced. This category included deaths from such different causes as birth defects and starvation due to maternal neglect. However, these deaths had in common that they resulted from Hazard Rates and Causes of Death / 143 TABLE 11. Life Table for Female Animals Number of animals ~ Age interval (years) 0-0.25 0.25-0.5 0.5-0.75 0.75-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 10-11 11-12 12-13 13-14 14-15 15-16 16-17 17-18 18-19 19-20 20-21 21-22 22-23 23-24 24-25 25-26 26-27 27-28 28-29 29-30 Alive at beginning of interval 192 166 160 155 152 127 106 86 77 69 60 51 41 33 30 27 25 20 16 13 12 10 8 7 6 6 6 5 5 3 1 1 0 Censored within interval Exposed to risk of dying Died within interval Died from violence Died from infection 1 4 3 2 19 17 19 9 8 9 9 7 8 3 3 2 5 4 3 1 1 1 1 1 0 0 1 0 2 1 0 1 191.1 164.1 158.5 155 143.1 118 98.2 80.8 73.2 64.9 57.1 47.9 36.3 32 28.9 26.1 22.2 17.4 14.6 12.3 11.4 9.1 7.5 6.5 6 6 5.6 5 3.6 2.5 1 0.8 25 2 2 1 6 4 1 0 0 0 0 3 0 0 0 0 4.5" 2 1 0 4 1.5" 0 0 0 0 0 1.5" 0 0 0 0 8.5" 0 0 0 2 0.5" 1 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0.5" 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 "In these cases, no conclusive decision could be made as to one particular cause of death. the animals' apparent biological unfitness for postnatal life and, therefore, were summarized under the same heading. Further deviating from the classification of Steffen et al. [19881, the category "infection" included all kinds of infectious diseases as well as sepsis associated with severe injuries. For estimating survival functions and hazard rates, I have used the nonparametric kernel method. The most important feature of this method is that it does not assume a specific functional model, i.e., a particular pattern of age-related survival and mortality. This aspect seems to be essential when considering the large variability in survival characteristics of Old World monkeys [Gage & Dyke, 19881. A major problem faced in this study was the large number of animals removed from the group for the purpose of sale. Such a high degree of censoring seems typical for active primate breeding colonies [Dyke et al., 19861 and apparently 144 I Luder sunrival Hazard Rate - Female Survival -- Male Survival Female Hazard ....... Male Hazard Hazard Rate -All Causes Immaturity -v i Infection -_ Other Causes ....... Unknown Cause Hazard Rate - All Causes Immaturity - VilenCe Infection -- Other Causes ....... Unknown Cause Fig. 1. Age-specific survival and mortality. a represents overall survival and hazard rates for male and female animals; b and c depict hazard rates related to the categories of causes of death “immaturity,” “violence,” “infection,” as well as “other” and “unknown causes” for female (b) and male (c) animals. affected most previous investigations of mortality in captive monkeys [Gage & Dyke, 1988; Tigges et al., 19881. As a result, the numbers of living animals and deaths recorded in the present Hazard Rates a n d Causes of Death / 145 study at older ages were so low that hazard rates due to senescence could not be estimated reliably. Furthermore, it cannot be quite ruled out that in a zoo, only sickly and weak animals are culled for transfer. As a result, censoring could have been related to health and vigor of the monkeys. Yet, no confirmation for this hypothesis was obtained from the vital records, as they did not reveal antecedence of particularly many fights or illnesses in censored animals. In fact, a majority of culled monkeys were transferred for purposes of breeding, and observations in two colonies which had obtained such animals indicated that they were healthy and vigorous. The obtained survival and hazard rates up to the age of about 10-15 years in both male and female animals were in good agreement with those reported previously for rhesus monkeys, although housing conditions, management of the colonies, and standard animal care may have varied considerably [Rawlins et al., 1984; Dyke et al., 1986;, Gage & Dyke, 1988; Tigges et al., 19881. When compared with the standard survivorship developed by Gage and Dyke , the survival functions of the animals analyzed in this study dropped more steeply after birth, but subsequently exhibited flatter, almost linearly decreasing courses up to the ages of about 8 years in males and 17 years in females. Corresponding with these survival functions, hazard rates dropped markedly after birth. During adolescence and early adulthood, they remained low and almost unchanged in females, whereas in males, the curve exhibited a peak a t about 6 years. After the age of about 7 years in males and 14 years in females, hazard rates increased steeply. Thus, on the basis of age-specific mortality, two periods or stages can be identified. During the first period after birth characterized by the rapid drop in overall hazard rates, all categories of causes of death contributed similarly to overall mortality. However, while “violence,” “infection,” as well as “other” and “unknown causes” to some degree occurred throughout further life, deaths resulting from “immaturity” were recorded only during the early period after birth. The association with deaths due to “immaturity,” therefore, seems to discriminate the first stage of mortality and to justify the term “immature component” suggested by Gage and Dyke 119881. The second period of low overall mortality was determined largely by a high incidence of violent deaths. In both genders, the respective hazard rates attained peaks during the early reproductive period of life and thereafter decreased markedly. This pattern was particularly prominent in males, where a transient increase in overall mortality around 6 years of age apparently was due to a high incidence of violent deaths. It could be argued that the peaks in hazard rates resulted merely from low numbers of deaths. However, the obtained rates for violent deaths probably are underestimates, as severe injuries often were associated with sepsis, and consequently deaths were assigned equally to the categories “infection” and “violence,” although the violent event most likely was the primary cause of death. Furthermore, a similar, uneven distribution of violent deaths among males and females and among various age groups is also suggested by the recent results of long-term demographic monitoring of free-ranging langurs [Rajpurohit & Sommer, 19911. In this population, (1)males faced a considerably higher risk to die from violence than females; (2) deaths of infant animals were most frequently due to intraspecific killing, which occurred only very rarely in the colony examined here; (3) juvenile animals died particularly often from accidents; and (4) heavy wounding as a result of fights, in addition to accidents, appeared to be the most frequent causes of death in subadult and adult animals. Conceivably, the high incidence of violent deaths during adolescence and early adulthood might be linked to specific aspects of simian reproduction, such as emigration of males from their natal 146 I Luder groups and intermale competition for dominance in a group, in which females can get involved as well. Thus, violent deaths could well account for a transient increase in overall mortality during the early reproductive period of life. In fact, peaks in overall hazard rates around 8.5 and 6 years of age, respectively, can also be demonstrated for the female and, most prominent, the male rhesus monkeys of the Yerkes Regional Primate Research Center [Tigges et al., 19881. Such a course is incompatible with current thinking of residual mortality, as this component is assumed to be constant throughout life. In addition, such a course is neither compatible with current concepts of senescence. As is clearly indicated by the widely established use of the Gompertz and Weibull functions to mathematically model the senescent component, this is commonly conceived as a continuously increasing process which does not allow for peaks in hazard rate. Thus, violent deaths and, hence, the major proportion of second stage mortality do not seem to be related to senescence, but rather to extrinsic environmental factors. This does not imply that biological aging processes could not be initiated already during the second stage, as is suggested, for instance, by age-associated changes in joint mobility of rhesus monkeys [DeRousseau et al., 19831. If so, however, the expression of biological aging is not revealed by mortality, but rather masked by the extrinsic factors prevailing during early adulthood. Therefore, when considering overall hazard rates in monkeys, senescence as the main cause of death should be assumed only after the end of the second stage. Although the present data do not allow me to estimate precisely the age of this end, they do suggest that in crabeating monkeys, it occurs between 10 and 15 years. This estimate is in good agreement with that suggested by the appearance of degenerative joint disease in rhesus monkeys [DeRousseau, 19851. However, it assumes the onset of manifestations of senescence clearly later than at puberty, as was done in a recent comparative analysis [Finch et al., 19901. CONCLUSIONS These findings would allow me to hypothesize that: (1) in a population of captive nonhuman primates, the period of mortality associated primarily with senescence is preceded by two other distinct stages; (2) while the early period following birth is characterized by deaths resulting from apparent unfitness for postnatal life, the second stage seems to be related mainly to extrinsic environmental factors; and (3) when considering overall hazard rates, the onset of manifestations of senescence in Macaca fascicularis should be assumed between 10 and 15 years of age. ACKNOWLEDGMENTS I am indepted to Mr. F. Salz and Dr. E. Thommen, Zoo of Basel, Basel (Switzerland) for providing life history records of the monkeys. Heartfelt thanks are due to Mrs. R. Kroni for preparing the manuscript. This study was supported by grants of the Hartmann-Muller Foundation (298 and 3081, and the Foundation for Dental Research of the Swiss Society of Dentists (144). REFERENCES DeRousseau, C.J. Aging in the musculoskeleta1 svstem of rhesus monkeys: 11. Degeneratiie joint disease. AMERiCAN JOaRNAL OF PHYSICAL ANTHROPOLOGY 67:177-184, 1985. DeRousseau, C.J.; Rawlins, R.G.;Denlinger, J.L. Aging in the musculoskeletal system of rhesus monkeys: I. Passive joint excursion. AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 61:483-494, 1983. Dyke, B.; Gage, T.B.; Mamelka, P.M.; Goy, R.W.; Stone, W.H. A demographic analysis of the Wisconsin Regional Primate Center Hazard Rates and Causes of Death / 147 rhesus colony, 1962-1982. AMERICAN JOURNAL OF PRIMATOLOGY 10:257269, 1986. Finch, C.E.; Pike, M.C.; Witten, M. Slow mortality rate accelerations during aging in some animals approximate that of humans. SCIENCE 249:902-905, 1990. Gage, T.B.; Dyke, B. Model life tables for the larger Old World monkeys. AMERICAN JOURNAL OF PRIMATOLOGY 161305320, 1988. Gasser, T.; Miiller, H.-G. Kernel estimation of regression functions. Pp. 23-68 in SMOOTHING TECHNIQUES FOR CURVE ESTIMATION. T. Gasser, M. Rosenblatt eds. Berlin, Heidelberg, New York, Springer Verlag, 1979. Miller, R.G. J r . SURVIVAL ANALYSIS. New York, John Wiley & Sons, 1981. Miiller, H.-G.; Wang, J.-L. Nonparametric analysis of changes in hazard rates for censored survival data: An alternative to change-point models. BIOMETRIKA 77: 305-314, 1990. Rajpurohit, L.S.; Sommer, V. Sex differences in mortality among langurs (Presbytis entellus) of Jodhpur, Rajasthan. FOLIA PRIMATOLOGICA 56:17-27, 1991. Rawlins, R.G.; Kessler, M.J.; Turnquist, J.E. Reproductive performance, population dynamics and anthropometrics of the freeranging Cay0 Santiago rhesus macaques. JOURNAL OF MEDICAL PRIMATOLOGY 13:247-259, 1984. Smith, D.G. A comparison of the demographic structure and growth of free-ranging and captive groups of rhesus monkeys (Macaca mulatta). PRIMATES 23:24-30, 1982. Steffen, H.; Cipriani, J.; Oglesby, S.;Wyrsch, R.; Bergauer, G. Swiss table of mortality. STATISTICAL REPORTS, Federal Office of Statistics, Berne, Switzerland, 1988. Tigges, J.; Gordon, T.P.; McClure, H.M.; Hall, E.C.; Peters, A. Survival rate and life span of rhesus monkeys a t the Yerkes Regional Primate Research Center. AMERICAN JOURNAL OF PRIMATOLOGY 15: 263-273, 1988.