Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Curr Anthropol. Author manuscript; available in PMC 2010 October 1.
Published in final edited form as:
Curr Anthropol. 2009 October; 50(5): 649–655.
PMCID: PMC2909026

What Do We Know About the Agricultural Demographic Transition?


The Agricultural Revolution accompanied, either as a cause or an effect, important changes in human demographic systems. This paper examines the evidence available for investigating the effects that the agricultural transition had upon human population growth, mortality, fertility, and health. The goal is to determine whether the existing data support the consensus model that compared to hunter-gatherers, fertility and mortality increased and health declined with the adoption of agriculture. Analysis of the agricultural transition relies primarily upon archaeological and paleodemographic data, and is thus subject to the errors associated with such data. To use these existing data, certain assumptions must be made, and these can profoundly affect the inferences that are drawn. While it is clear that, in general, population growth accompanied the agricultural transition, it is not as clear exactly how fertility and mortality changed, nor whether the transition caused a decline in health. Though the model of the agricultural demographic transition as outlined here may be correct, we urge readers to remain aware of the underlying assumptions and to be open to future empirical evidence.

Although often called a theory, the “Demographic Transition” (Thompson, 1929) is an empirical model of the growth of human populations with industrialization. Population growth is considered a cause, in the Bosrupian view, or an effect, in the Malthusian view, or possibly both, of the industrial revolution. In any event, the proximate cause of human population growth over the industrial era was/is due to a decline in mortality beginning in Western Europe in the late 18th and early 19th centuries, followed some years later by a decline in fertility. Populations grew, or at least the rate of growth accelerated during the period between the decline in mortality and the decline in fertility. Of course, the Demographic Transition is a “model”, i.e. a simplification of reality. There is variation from population to population in terms of the time of onset, speed of the decline in mortality and fertility, and the length of delay between the two, etc. Further, the industrial Demographic Transition is not complete. We do not yet know if the postindustrial populations will return to a stationary state with low mortality and fertility.

The “Demographic Transition” has been elaborated by the “Epidemiological Transition” (Omran, 1977) and the “Health Transition” (Riley, 1989). The Epidemiological Transition is an empirical model that describes consistent shifts in cause of death that accompany the Demographic Transition. In particular, this model argues that the frequency of the infectious causes of death declines, while the frequency (not necessarily the risk (Gage, 2005)) of degenerative causes of death increases. The “Health Transition” adds a description of the secular trends in morbid states accompanying the Demographic and Epidemiologic Transitions. This model summarizes the general finding that the prevalence (not necessarily the incidence) of morbidity increases as the Demographic Transition progresses(Riley, 1989; Roos, Havens, & Black, 1993). In particular, there appears to be a generally inverse relationship between mortality and morbidity. Perhaps the most convincing evidence concerning the inverse relationship of mortality and morbidity is not the general increase in morbidity observed with the decline in mortality, but rather that morbidity has declined in instances where mortality has experienced secular increases since the advent of the industrial revolution (for example, areas of Eastern Europe after the breakup of the Soviet Union) (Riley, 1992).

The demographic and health transitions illustrate the fact that mortality and health are not directly related (Gage, 1989; Usher, 2000; Crimmins, Hayward, & Saito, 1994). This is because mortality is an “absorbing state”—it removes people from the population. If people in poor health are more likely to die than people in good health, there will tend to be an inverse relationship between the level of health and mortality as observed by Riley (1992). Again, this is a simplification, a model of reality. The relationship between health and mortality is likely to be complex (Crimmins, et al.,1994; Crimmins & Saito, 2001) and depends critically on one's definition of health. Death is fairly well-defined, physiologically. Health, on the other hand, has many definitions and might best be considered a multifactorial condition. In medicine, health is traditionally defined as the absence of disease; the World Health Organization has expanded this definition to include “complete physical, mental, and social well-being”. The definition typically used in descriptions of the Health Transition is the ability to perform “activities of daily life”, (e.g., dressing and feeding one's self). Increasingly, more “rigorous” physiological measures are used as biodemographic measures are incorporated into large health surveys. On the other hand, the concept of “activities of daily life” as a general measure of health should be particularly appealing to anthropologists. It is an attempt to measure the frequency with which members of the population can carryout the tasks that need to be done for individual and group survival (Thomas, Gage, & Little, 1989). Note that this is culturally relative. The activities of daily life for a hunter-gatherer, an agriculturalist, or an industrialist differ radically, and their relationship to individual physiological capabilities may differ as well.

It is clear that a reorganization of human demographic systems similar to the industrial Demographic Transition occurred prehistorically either as a cause or an effect of the Agricultural Revolution. The general (consensus) model suggests that among hunter-gatherers, mortality was moderate and fertility relatively low (due perhaps to lactational amenorrhea); fertility and mortality increased with the adoption of agriculture, while “health” declined. This model is based on the concept that fertility increased with agriculture (that is, the low forager fertility hypothesis) and that population growth was 0.0 or at least small among hunter-gatherers and post-transition agriculturalists, a Malthusian assumption. If so then mortality must have increased, at least among post-transition agriculturalists. The exact dynamics of mortality and fertility during the transition is unknown. However, populations grew because fertility exceeded mortality at some point during the transition (Buikstra, Konigsberg, & Bullington, 1986; Howell, 1986; Bocquet-Appel 2002; Bocquet-Appel, & Naji 2006; Bocquet-Appel 2007). Health is considered to decline because the proportion of skeletons with pathologies increased after the Agricultural Revolution (Cohen, 1989; Cohen & Armelagos, 1984). Note that this view when combined with the agricultural demographic transition varies from the theoretically inverse association of mortality and health, which is commonly observed during the industrial demographic transition.

The industrial and agricultural demographic transitions differ in two respects. First, when investigators consider the industrial demographic transition they are referring to the dynamics during the transition with the knowledge that the transition is (or may be) ongoing. When investigators consider the agricultural transition they often contrast pre-versus post-transition demographic regimes, rather than examine the demographic trends during the transition itself. Second, demographic transitions are intended to be empirical (descriptive) models of the secular trends in mortality, fertility, population growth, and health; consequently, construction of such a model is relatively simple when the data are directly observable. However, demographic data for the prehistoric period are not directly observable and typically must be estimated based on a set of theoretical assumptions. These issues of formal paleodemography are not new and have been repeatedly discussed in the literature (Moore, Swedlund, & Armelagos, 1975; Wood, Milner, Harpending, & Weiss, 1992). Unfortunately, these issues are often ignored or superficially acknowledged and then ignored. The aim of this paper is to review what we currently know about the agricultural demographic transition in light of these issues.

Population Growth

Perhaps the most reliable data concerning the agricultural transition are changes in population size during the transition to agriculture. In general, populations grew, although there are no doubt exceptions. The empirical evidence consists largely of archeological survey data (Hassen, 1979). Whether population growth was 0.0 pre- and post- transition is not as clear. The assumption that mortality increased during the agricultural transition depends upon this Malthusian assumption. Von Foerster, Mora, and Amiot (1960) have modeled human population growth from prehistoric times through the 1960s and shown that human population growth has been faster than exponential, i.e. it has accelerated with time. This is at variance with the assumption of 0.0 population growth, although it is possible that local populations experienced Malthusian growth dynamics, but that agriculture and later industrialization spread at ever-increasing rates so that the total human population expanded at an ever-increasing rate. It is sobering to note that Von Forester predicted human extinction in 2020 when his model predicts that the human population will reach infinity. While this seems patently ridiculous, United Nations estimates of world population levels only fell below Von Forester's modeled predictions circa 1995 (Gage unpublished calculations).


Mortality is typically measured using a life table approach to estimate expectation of life. There are several issues associated with paleodemographic applications of the life table: taphonomy, age estimation, and the assumption of stationarity (0.0 population growth). Paleodemographic samples are potentially unrepresentative because of the variable taphonomic processes of burial, preservation, and recovery. This will not be discussed further here except to note that if the biases are quantified (Walker, Johnson, & Lambert, 1988), they can be included in demographic estimation.

The problems of age estimation using skeletal data have been widely discussed and may have recently been resolved. Briefly, the problem is that standard methods of estimating age using known-age reference collections bias the estimated ages toward the age structures of the reference collections. This phenomenon was first pointed out by Bocquet-Appel and Masset (1982). However, only recently has a potential cause of this bias been identified and Baysian methods implemented to correct it (see Hoppa & Vaupel, 2002 for discussion of what is now known as the “Rostock Manifesto”). It is too early to tell if these new age estimation methods will revolutionize paleodemographic life tables. For example, will they result in age-specific mortality curves for paleodemographic lifetables that more closely resemble those of historic populations, including contemporary hunter-gatherers (Gage, 1989)? Further, will they increase the range of expectations of life for paleodemographic life tables, which currently all fall within a rather restricted range, approximately 20–30 years at birth (Gage, 2000)?

The assumption of stationarity has also been widely discussed (Moore, et al., 1975; Wood et al., 1992). The assumption that a population is stationary (closed to migration and having constant age-specific fertility and mortality rates, a stable age distribution, and a growth rate of zero) is useful for paleodemography because it allows one to estimate age-specific mortality rates and other life table parameters from skeletal age-at-death distributions. However, if a population is not stationary—if, for example, it has a non-zero growth rate—the estimated life table can be profoundly distorted. Researchers have shown, perhaps counterintuitively, that the age-at-death distributions of non-stationary populations are more strongly correlated with changes in fertility rates than changes in mortality rates. For example, if a population experiences growth, which we expect to be the case during the transition to agriculture, there will be an increase in the proportion of young individuals in the population, and expectation of life will be underestimated (regardless of the causes of growth). If, on the other hand, a population declines, expectation of life will be overestimated. In fact, under the stationary assumption, if populations grew during the transition to agriculture, the only unambiguous signal of a change in mortality would be an increase in expectation of life (mean age at death) (Eshed, Gopher, Gage, & Hershkovitz, 2004), which would indicate that mortality declined and fertility did not increase sufficiently to obscure the shift in mortality. On the other hand, evidence of a decrease in expectation of life could be due to a positive growth rate rather than a true decline in expectation of life. The situation is reversed if the growth rate of a population is negative. Consequently, under the stationary assumption, mean age at death is not a good method of estimating trends in mortality during the transition to agriculture.

A less stringent assumption is that the population under consideration was stable—that is, all the assumptions of the stationary population except for that of zero population growth. Stability is potentially a serious problem when dealing with an age structure at a particular point in time, specifically for demographers observing living populations. However, archaeological age-at-death assemblages are deposited over an extended period, and all populations oscillate around their stable states. Consequently, paleodemographic assemblages as averages across a number of years may be close to stable conditions even though the population at a particular point in time is not. Very recently, a method of simultaneously estimating r (the intrinsic growth rate) and a life table from age-at-death distributions has been developed (Holman, personal communication, 2007), which is also compatible with the “Rostock Manifesto”. Finally, some have argued that even the assumption of stability may be flawed when examining paleodemographic assemblages. Bonneuil, (2005) has recently suggested a possible solution to this problem. These recent developments may revolutionize our understanding of paleodemography. However, to our knowledge they have not yet been practically applied. To date, most paleodemographic mortality estimates assume the population is stationary.

If it can be assumed that population growth fluctuated around r = 0.0, which is often done (Cohen this volume), then averaging a series of paleodemographic life tables might provide more reasonable results than a single paleodemographic life table. Averaging rates across 29 paleodemographic life tables indicates that hunter-gatherers have a mean expectation of life of 21.6 years; horticulturalists, a mean of 21.2 years; and agriculturalists, 24.9 years. None of these differences are statistically significant (Gage, 2000), thus these data do not support the hypothesis that the mortality of agriculturalists was higher or lower than hunter-gatherers.


Sattenspeil and Harpending (1983) have shown that mean age-at-death (expectation of life if a population is stationary) is a better indicator of fertility than mortality, at least in stable populations. Consequently, it has become popular to use the changes in mean age-at-death of a skeletal sample or ratios of several broad ages at death (Bocquet, 1979; Buikstra, et al., 1986) to indicate changes in fertility rather then changes in mortality. While this seems counterintuitive, the original (Sattenspiel & Harpending, 1983) method works reasonably well under certain conditions (Horowitz, Armelagos, & Wachter, 1988), specifically, Coale and Demeny Model West Mortality at low expectations of life, and low growth rates. It is unclear if these assumptions are correct. The empirical evidence suggests low expectations of life (20–30 years), but as noted above these empirical estimates are flawed. Similarly, Coale and Demeny Model West Mortality is frequently assumed (whenever better information is unavailable) because it is “average” for industrial era populations (Coale, Demeny, & Vaughan, 1983; Gage, 1990)). In this regard, the Coale and Demeny model life tables are not based on any empirical life tables with expectations of life at birth lower than 35 years. Life tables available to them with expectations of life below 35 years were considered flawed (and eliminated from the analysis) because they displayed unusual age patterns of mortality. The model life tables with expectations of life between 20 and 35 in the Coale and Demeny model life tables are extrapolated from regressions on life tables with expectations of life of 35 years and above. The empirical paleodemographic life tables are clearly divergent from all the Coale and Demeny model life tables but most closely resemble the Coale and Demeny South model life table, not the West model (Gage, 1990). Horowitz, et al. (1988) conclude that like mortality, good estimates of the birth rate also require an estimate of r.

How robust are the ratio approaches of estimating fertility from the real (and currently unobserved) variation in the shape of the human mortality curve? There is no doubt that the death ratios (Bocquet-Appel, 1979; 2002; Bocquet-Appel & Naji, 2006; Buikstra, et al., 1986) pick up consistent albeit weak signals indicating a change in demographic regime during the transition to agriculture. It is even likely that this is due to changes in fertility, as ethnographic analogies suggest that the fertility of sedentary agriculturalists is higher than forager fertility (Bently, Jasienska, & Goldberg, 1993; Hewlett, 1991)). The mean completed fertility for a sample of 47 populations of various economies is 6.2 children with a range from 3.0 to 10.0. The mean fertility for a) hunter-gatherers is 5.6 with a range of 3.5 to 8.0, b) horticulturalists is 5.4 with a range of 3.0 to 7.0, and c) agriculturalists is 6.6 with a range of 3.5 to 10.0. Statistical examination among the subgroupings suggests no differences in the level of fertility between hunter-gatherers and horticulturalists, but supports differences between agriculturalists and both hunter-gatherers and horticulturalists. However, other possible explanations for the death ratio results remain, such as age specific changes in mortality.


Skeletal samples have been used as a source of data regarding the health consequences of changes in subsistence strategies. Such samples provide data on skeletal health (i.e., the presence or absence of skeletal lesions or stress-markers), which is interpreted in terms of general health. Several researchers have observed an increase in the frequency of certain skeletal lesions or stress-markers in samples from agricultural populations compared to hunter-gatherers, and they view these changes in apparent skeletal health as persuasive evidence of a decline in general health at the transition to and/or intensification of agriculture. Many of the contributors to Cohen and Armelagos' (1984) seminal volume present evidence of increases in frequencies of skeletal lesions and a decrease in stature in agricultural samples compared to hunter-gatherers. Despite caveats from some contributors regarding small sample sizes and the lack of clear patterns with respect to certain skeletal stress-markers, the editors argue that there is sufficient evidence to conclude that the adoption of agriculture was generally detrimental to human health and quality of life. Since the Cohen and Armelagos volume, many researchers have continued to find similar evidence of a decline in skeletal health associated with agriculture.

Until relatively recently, most good data on the effect of the transition to agriculture came from North American sites (Cohen, 1989; Larsen, 2002), and little was known about the consequences of the transition for much of Europe, Africa, Asia, and South America. This is changing, as paleopathologists widen their geographic focus; several studies summarized in the Cohen and Crane-Kramer volume (2007) provide evidence from these neglected areas, much of which confirms the above pattern of declining skeletal health.

Some researchers question what skeletal health really tells us about general health conditions. As mentioned above, general health can be defined in several ways. If we define health using the standard medical definition of an absence of disease, then an increase in the prevalence of skeletal lesions (i.e., a decline in skeletal health) might indicate a decline in general health. However, such changes in lesion frequencies are only capturing changes in diseases that cause skeletal lesions, and important determinants of health and mortality in past populations are potentially undetected using skeletal samples; for example, according to Cook (2007), in pre-modern populations, the most common causes of death were pneumonia and gastrointestinal infections, which do not cause bone lesions. If, rather, we define health as the ability to perform the activities of daily life, concluding that an increase in skeletal lesions indicates a decline in health assumes that those lesions are associated with factors that reduce an individual's ability to perform daily activities. But, is this assumption always valid? If health is defined as a lack of conditions that are known to increase an individual's risk of death, a decline in the general health of a population would be indicated by a decline in skeletal health accompanying a decrease in life expectancy. To examine temporal trends in health, Steckel and Rose (2002) developed a health index that incorporates information about the presence of several skeletal lesions and individual age-at-death; age-specific rates of lesions are weighted by the distribution of person-years lived by age in a Model West level 4 reference population. According to Steckel and Rose (2002), comparing the health index with the estimated life expectancy for a particular sample can reveal whether the prevalence of lesions is associated with good or poor health, thereby addressing one aspect of the “osteological paradox” (Wood, et al., 1992).

Some researchers have questioned whether poor skeletal health is always necessarily a sign of declining general health within a population. The perhaps counterintuitive argument is that in some contexts, increases in the frequency of certain lesions might indicate general improvements in health (Ortner, 1991; Wood et al., 1992). This argument is based on the fact that visible skeletal lesions take some time to form; they do not form immediately in response to trauma or disease, but rather take weeks or months to become detectable. Individuals with skeletal lesions might therefore have been healthier than their peers without lesions, given that they were able to survive malnutrition, trauma, or disease long enough for the skeletal lesions to form. Absence of a certain skeletal lesion might indicate relatively poor health, as individuals without lesions were in such poor health that they succumbed to illness, trauma, or malnutrition and died before lesions ever formed, consistent with the observed trends during the industrial demographic transition.

Cohen and Crane-Kramer (2007) argue that concerns about potential paradoxes in skeletal samples are unwarranted, as multiple lines of evidence, including paleopathological and ethnographic data, often match theoretical expectations. Similarly, Steckel and colleagues (2002) claim a positive correlation between skeletal health and expectation of life in paleodemographic data, although these data are based on biased age-estimation methods and expectation of life is based on guesstimates of population growth. Using a multi-state model developed for paleodemographic studies by Usher (2000) and the “Rostock Manifesto” for age estimation, DeWitte and Wood (2008) found that skeletal lesions were associated with increased risks of death in both normal and catastrophic mortality samples. However, not all researchers are explicit about the relationship between the lesions they observe and health (however they define it). We do not deny that there is a general trend of increasing prevalence of skeletal lesions with the transition to agriculture; however, there are important links between the presence of certain lesions and their effect on survival and/or health that remain to be clarified.


Given what we know, how then do we arrive at the model of the agricultural demographic transition outlined above? We know that population size and density increased during the transition to agriculture. This is based on empirical archeological evidence and secondarily inferred from genetic evidence of population expansions dated to the transition to agriculture. We then assume that fertility increased with the advent of (at least) intensive agriculture based on ethnographic analogies, and inferred from the ratio methods of fertility estimation discussed above, although contemporary horticulturalists (incipient agriculturalists?) appear to have low fertility. If it is further assumed that population growth before and after the agricultural transition is essentially 0.0 (a Malthusian assumption and contrary to von Foerster's analysis), then mortality must have increased among post-agricultural populations. Of course, if we accept the hypothesis that growth before and after the agricultural transition was essentially 0.0, then we can accept the paleodemographic mortality estimates as well, up to the stable assumption. These estimates are not consistent with higher mortality among even intensive agriculturalists. On the other hand, if human population growth has increased at an accelerating rate through time as suggested by von Foerster's analysis and Bosrupian theory, then it is possible that fertility increased, and mortality decreased with the agricultural transition. If mortality decreased and the negative relationship between mortality and health is correct, then health might have declined as suggested by paleopathological data. Note this model is not a simple empirical model like the industrial demographic transition. Current models of the agricultural demographic transition are largely based on assumption. We do not reject the description of the agricultural demographic transition as generally outlined. Nor do we wish to declare a farewell to paleodemography. We do want to point out that this scenario is based on a series of assumptions that are all questionable (at best). We hope that these preconceived notions of the agricultural demographic transition do not delay acceptance of more rigourous empirical evidence when it becomes available. Given the recent advances in “formal paleodemography” we expect that this evidence may become available soon.

figure 1
Figure one presents the “consensus” view of the Demographic Transition that accompanied the transition to agriculture under the Malthusian assumption that growth was zero before and after the transition. The heavy solid line represents ...


  • Bently GR, Jasienska G, Goldberg T. Is the fertility of agriculturalists higher than that of nonagriculturalists? Current Anthropology. 1993;34:778–85.
  • Bocquet JP. Une approche de la fécondité des populations inhumées. Bulletins et Mémoires de la Société d'Anthropologie de Paris. 1979;6:261–68.
  • Bocquet-Appel JP. Paleoanthropological traces of a Neolithic Demographic Transition. Current Anthropology. 2002;43:637–650.
  • Bocquet-Appel JP. Testing the hypothesis of a worldwide Neolithic Demographic Transition. Current Anthropology. 2006;47:341–365.
  • Bocquet-Appel JP, Masset C. Farewell to paleodemography. Journal of Human Evolution. 1982;11:321–33.
  • Buikstra JE, Konigsberg LW, Bullington J. Fertility and the development of agriculture in the prehistoric Midwest. American Antiquity. 1986;51:528–46.
  • Coale AJ, Demeny P, Vaughan B. Regional Model Life Tables and Stable Populations. Academic Press; New York: 1983.
  • Cohen MN. Health and the Rise of Civilization. Yale University Press; New Haven: 1989.
  • Cohen MN, Armelagos GJ, editors. Paleo-pathology at the Origins of Agriculture. Academic Press; Orlando: 1984.
  • Cohen MN, Crane-Kramer GMH. Ancient Health: Skeletal Indicators of Agricultural and Economic Intensification. University Press of Florida; Gainesville, FL: 2007.
  • Cook DC. Maize and Mississippians in the American Midwest: Twenty years later. In: Cohen MN, Crane-Kramer GMH, editors. Ancient Health: Skeletal Indicators of Agricultural and Economic Intensification. University Press of Florida; Gainesville, FL: 2007. pp. 10–19. 2007.
  • Crimmins EM, Hayward MD, Saito Y. Changing mortality and morbidity rates and the health status and life expectancy of the older population. Demography. 1994;31:159–75. [PubMed]
  • Crimmins EM, Saito Y. Trends in healthy life expectancy in the United States, 1970–1990: Gender, racial, and educational differences. Social Science and Medicine. 2001;52:1629–41. [PubMed]
  • DeWitte SN, Wood JW. Selectivity of the black death with respect to preexisting health. PNAS. 2008;105:1436–41. [PubMed]
  • Eshed V, Gopher A, Gage TB, Hershkovitz I. Has the transition to agriculture reshaped the demographic structure of prehistoric populations? American Journal of Physical Anthropology. 2004;124:315–29. [PubMed]
  • Gage TB. The bio-mathematical approaches to the study of mortality. Annual Yearbook of Physical Anthropology. 1989;32:185–214.
  • Gage TB. Variation and classification of human age patterns of mortality: Analysis using competing hazards models. Hum. Bio. 1990;62:589–617. [PubMed]
  • Gage TB. Demography. In: Stinson S, Bogin B, Huss-Ashmore R, O'Rourke D, editors. An Evolutionary and Biocultural Approach. John Wiley and Sons; New York: 2000. pp. 507–552.
  • Gage TB. Are modern environments really bad for us?: Revisiting the demographic and epidemiological transitions. Yearbook of Physical Anthropology. 2005;48:96–117. [PubMed]
  • Hassen FA. Demographic anthropology. Annual Review of Anthropology. 1979;8:137–60.
  • Hewlett BS. Demography and childcare in preindustrial societies. Journal of Anthropological Research. 1991;47:1–37. [PubMed]
  • Hoppa RD, Vaupel JW, editors. Cambridge Studies in Biological and Evolutionary Anthropology. Vol. 31. Cambridge University Press; Cambridge: 2002. Paleodemography Age Distributions from Skeletal Samples.
  • Horowitz S, Armelagos GJ, Wachter K. On generating birth rates from skeletal populations. American Journal of Physical Anthropology. 1988;76:189–96. [PubMed]
  • Howell N. Demographic anthropology. Annual Review of Anthropology. 1986;15:219–46. [PubMed]
  • Larsen CS. Post-Pleistocene human evolution: Bioarchaeology of the agricultural transition. In: Ungar PS, Teaford MF, editors. Human Diet: its Origin and Evolution. Bergin and Garvey; Westport, Connecticut: 2002. pp. 19–36.
  • Moore JA, Swedlund AC, Armelagos GJ. The use of life tables in paleodemography. In: Swedlund AC, editor. Population Studies in Archaeology and Biological Anthropology: A Symposium. Vol. 30. Society for American Archaeology; Washington, D.C.: 1975. pp. 57–70.
  • Omran AR. Epidemiological transition in the U.S. Population Bulletin. 1977;32:3–42.
  • Ortner DJ. Theoretical and methodological issues in paleopathology. In: Ortner DJ, Aufderheide AC, editors. Human paleopathology: current syntheses and future options. Smithsonian Institution Press.; Washington, DC: 1991. pp. 5–11.
  • Riley JC. Long-term morbidity and mortality trends: Inverse health transitions. In: Caldwell JC, Findley S, Caldwell P, Santow G, Cosford W, Braid J, Broers-Freeman D, editors. What We Know About the Health Transition: The Cultural, Social and Behavioral Determinants of Health. Australian National University; Canberra: 1989.
  • Riley JC. From a high mortality regime to a high morbidity regime: Is culture everything in sickness? Health Transition Review. 1992;2:71–87. [PubMed]
  • Roos NP, Havens B, Black C. Living longer but doing worse: Assessing health status in elderly persons at two points in time in Manitoba, Canada, 1971 and 1983. Social Science and Medicine. 1993;36:273–82. [PubMed]
  • Sattenspiel L, Harpending H. Stable populations and skeletal age. American Antiquity. 1983;48:489–98.
  • Steckel RH, Rose JC, editors. The Backbone of History: Health and Nutrition in the Western Hemisphere. Cambridge University Press; New York: 2002.
  • Thomas RB, Gage TB, Little M. Reflections on adaptive and ecological models. In: Little M, Haas JD, editors. Human Population Biology: A Transdisciplinary Science. Oxford University Press; Oxford: 1989. pp. 296–319.
  • Thompson W. Population. American Journal of Sociology. 1929;34:959–75.
  • Usher BM. A multistate model of health and mortality for paleodemography: Tirup cemetery [Unpublished dissertation] Pennsylvania State University; 2000. 2000.
  • von Foerster H, Mora PM, Amiot LW. Doomsday: Friday, 13 November, A.D. 2026. Science. 1960;132:1291–307. [PubMed]
  • Walker PL, Johnson JR, Lambert PM. Age and sex biases in the preservation of human skeletal remains. American Journal of Physical Anthropology. 1988;76:183–88. [PubMed]
  • Wood JW, Milner GR, Harpending HC, Weiss KM. The osteological paradox: Problems of inferring prehistoric health from skeletal samples. Current Anthropology. 1992;33:343–70.