The Evolution of Human Longevity
from the Mesolithic to the Middle Ages:
An Analysis Based on Skeletal Data

by Jesper L. Boldsen & Richard R. Paine

The mythology concerning human longevity in the past is extensive. The first ten men mentioned in the Bible, Adam through Noah (except Cain and Abel), lived on average more than 850 years (Genesis 5: 1-32). Methuselah lived for 969 years. This was the maximum, but by no means exceptional. However, already in antediluvian time God fixed the living day of man to 120 years. Apparently, it took some time to adjust to this standard. The mean longevity of the eight male generations from Noah to Abraham was close to 300 years. There was a clear and statistically significant tendency to decline (r = -0.88, t = 4.58, df = 6, p < 0.001; Genesis, 11:10-26).

        Until recently it was believed that people in the European Middle Ages, if they survived the hazards of infancy and childhood, lived 'as long as people do today'. Only palaeodemographic studies of large samples of skeletons representing the general population can contribute to the thorough rejection of such false assumptions (Gejvall 1960, Boldsen and Vellev 1981, Boldsen 1984 and 1988, and Boldsen, Kieffer-Olsen and Pentz 1985). However, it is only possible to conduct thorough palaeodemographic studies on recently excavated skeletons and on material from the Middle Ages. In spite of this, this chapter tries to recover information on longevity in the more distant past by utilizing skeletal material covering most of European postglacial history. In doing so we will have to make sweeping generalizations and aim at pitifully simplistic conclusions about the demography of past communities.

        The original title of this chapter was: Extreme Longevity from the Mesolithic to the Middle Ages. However, there is nothing extreme about the old ages discussed in this chapter. In fact, determination of skeletal age at death is rather inaccurate in all but the youngest adults. It is only possible with available methods to estimate the fraction of the population that lived into a given anatomically defined age-related stage. This chapter presents an overview of changes of the fraction of the adult population that reached such a stage. The time period covered comprises some 8,000 years from the Mesolithic to the late Middle Ages.

Temporal frame, Material and Methods

Many students of human longevity might not be familiar with the sequence and classification of prehistoric periods in Europe. Consequently, the outlines of a temporal frame will be provided. It is hoped that this frame will help guide the reader's thoughts along the lines of the argument. The datings are in Table 1. As a biological species man was created in the Palaeolithic - a period stretching from around 2.5 million years ago to the end of the last glaciation at about 8300 BC.

        Man has not remained genetically unaltered through the last 10,000 years or so, but the most important changes in this period have occurred in the socio-cultural sphere rather than in the genetic make up of our species.

        The end of the last Ice Age saw the start of a new era of European history - the Mesolithic. Both the Upper Palaeolithic and the Mesolithic periods were characterised by subsistence based on foraging - gathering and hunting. Right at the end of what was the Ice Age of central and northern Europe, people in the Middle East adopted a new way of living - agriculture. They underwent the so-called Neolithic revolution. Agriculture became the dominant way of subsistence in central Europe by 5000 BC. The transition in southern Scandinavia was delayed by some 1000 years (Jensen 1988). The Mesolithic skeletons were derived from six different sites from a wider area of Europe than the skeletons from other periods. In fact, three of the samples are from Scandinavian sites (Albrethsen and Brinch Petersen 1976 and Larsson 1989). There are two samples from France and only one sample derived from a site in central Europe (Saller 1962).

        The Neolithic period lasted several thousand years. Starting with the formation of small, probably only semi-sedentary tribal communities subsisting on slash and burn agriculture and a lot of gathering and hunting. It was in this period permanently socially stratified societies evolved in Europe. During the latter half of the Neolithic period the Indo-European language community and the cultures that went with it spread over the face of the continent (Mallory 1989). The boundary between the Late Stone Age and the Early Bronze Age is quite arbitrary. In Denmark it is actually defined by the appearance of special types of flint tools and not by bronze implements as such (eg Rasmussen 1990). Neolithic skeletal samples are derived from nine sites. One site from each of the countries Austria (Ehgartner 1959), Czechoslovakia (Buchvaldek and Koutecky 1970) and France (Patte 1971) and six sites from Germany (Bach 1978, Czarnetzki 1966 and Grimm 1958, 1959 and 1961).

        In the Bronze Age urban communities formed in central and southern Europe. However, the majority of the population went on subsisting on agriculture. Many culturally important events took place during the Bronze Age, eg the historical events behind the narratives in much of the Old Testament and the travels of Odysseus. Like the transition from the Neolithic period to the Bronze Age the beginning of the following period, the Iron Age, was a long and slow process. The Bronze Age skeletal data come from two Austrian samples (Bertemes 1989 and Berner, no date).

        Iron metallurgy was invented early in the second millennium BC. But many centuries were to pass before iron came into regular use in Europe. The Iron Age framed European classical history. The rise and fall of the Roman Empire took place during this period. But like in the Bronze Age the majority of the population went on living in small rural communities only contributing to the spectacular events of written Iron Age history by being soldiers and paying taxes. Iron age skeletons are treated in four categories: central European Pre-Roman Iron Age, central European Roman Iron Age, Danish Early Roman Iron Age and Danish Late Roman Iron Age. The central European Pre-Roman Iron Age skeletons originate from six German sites (Spindler 1977, Schwidetzky 1978, Ehrhardt and Simon 1971, Kramer 1964 and Keiling 1974). Two central European Roman Iron Age skeletal samples are included, from Germany (no reference) and Hungary (Ery 1973). Data on all Danish Iron Age skeletons have been derived from Sellevold, Hansen and Jørgensen (1984).

        The Middle Ages which followed the Iron Age was the period of universal Catholicism in central and northern Europe. In central and southern Europe the Middle Ages started at the conquest of Rome in AD 410. But in Scandinavia the Viking Age, generally considered a part of the Iron Age, lasted till around AD 1000 (Sawyer 1988). The urban communities grew during the Middle Ages and around AD 1500 the urban population constituted between 10 and 20 per cent. The Medieval skeletal samples are divided into three categories, central European Early Medieval skeletons, central European Late Medieval skeletons and Scandinavian Medieval skeletons. The central European Early Medieval skeletons were excavated at seven different sites, two in Austria (no reference and Szilvassy 1980), three in Germany (Schnurgein 1987, Neuffer-Müller 1983 and Grünewald 1988) and two in Hungary (Garam 1979 and Nemeskeri 1963). The Scandinavian skeletons originate from six different sites in what was Denmark in the Middle Ages. In fact three of them are situated in Scania which has been a part of Sweden since 1660 (Boldsen 1979 and 1984, Boldsen, Kieffer-Olsen, and Pentz 1985, Persson and Persson 1984, Tkocz and Brøndum 1985 and Mårtensson 1976 and 1981).

        All the central European skeletons and most of the Scandinavian skeletons have had their age at death determined by others than the authors. Most of the age categories used can be related to closure of cranial sutures. In the classical German typology three age related anatomical stages have been defined for skeletons of people living to over 18-20 years. These stages comprise adultus/adulta (young people with no external closure of the medial sagittal suture), maturus/matura (middle aged people with some but not complete closure of the medial sagittal suture) and senilis/senilia (older people with complete closure of the medial sagittal suture). No generally accepted ages of transition between these stages exist. In order to give an idea about the range of transition ages it has been estimated in the Danish Medieval Tirup sample that the median maximum-likelihood age estimate at transition from adultus to maturus is 32.5 years and that the standard deviation of this transition is 5.4 years. For the transition from maturus to senilis the corresponding figures are 48.9 and 7.2 years.

        Figure 1 illustrates the estimated probabilities of at least having reached the stages adultus/adulta, maturus/matura and senilis/senilia in the Tirup skeletal sample. As expected it appears that the transition into adultus/adulta is much less variable than the transition into maturus/matura and senilis/senilia. This is in accord with the general finding that age estimation becomes increasingly inaccurate with increasing age at death. It can be assumed that most females through the studied period only were fertile during adulta.

Results and Discussion

Figure 2 summarizes the development of survival to maturus/matura for the two sexes conditional on survival to around age 18-20.

        Females experienced a very serious drop in the probability of reaching matura from the Mesolithic to the subsequent periods. This drop was followed by an increase from Pre-Roman Iron Age and onward. In males nothing seemed to change in the probability of reaching maturus from the Mesolithic to the Pre-Roman Iron Age. The sharp increase in the chance of reaching maturus in males observed in the Roman period is probably a consequence of sample bias. This spike is only based on 191 skeletons from two badly described sites in central Europe.

        Cultural evolution was not synchronised in Europe from the Mesolithic to the Middle Ages. Central Europe was usually far in advance of southern Scandinavia. Perhaps Mesolithic foraging was different in coastal Scandinavia and continental central Europe. This might not have had any great demographic consequences; but from the Neolithic revolution the two regions followed different trajectories of cultural evolution dependent on the local environmental conditions, the regional cultural dynamics and impulses received from the rest of the World.

        Figure 3 illustrates the development of the chance of reaching maturus/matura in males and females from southern Scandinavia from the Early Roman Iron Age to the Middle Ages. The structure of this development is quite different from the structure of the development in central Europe.

        In Scandinavia, it appears that the chance of reaching maturus/matura declined from the Early Roman Iron Age to the Middle Ages. However, this is probably due to sampling bias. The Danish Roman Iron Age skeletal sample is heavily dominated by the upper classes, the regional or national élite of those days. The Medieval samples on the other hand have been selected to represent the general population in various parts of the region. This interpretation is supported by the fact that in all three periods survivorship is greater in the Scandinavian series than in the central European series.

        A preliminary summary conclusion about the evolution of the patterns of adult mortality over the last 8-9000 years is summarized in Figure 4.

        Figure 4 is based primarily on the data presented in Figure 2. However, the points for Roman Iron Age and the late Middle Ages have been omitted and the latter have been replaced with much larger and much better described data from southern Scandinavia. As an illustration of the dramatic demographic changes in recent centuries the survivorship to age 40 given survivorship to age 20 is shown for contemporary Denmark. Figure 4 shows a u-shaped curve for the female chance of surviving to matura and a constant curve of the male chance with an increase starting in the Iron Ages. Great differences in the level of survivorship among the sexes persisted well into the second millennium AD.

        Three of the four urban samples from Medieval southern Scandinavia show no significant female adult surplus mortality. Only one urban sample (Sct. Mikkel - situated just outside the city wall) shows a significant female surplus mortality. On the other hand, both the two large rural samples do show a highly significant female surplus mortality. It is worth noting that the structure of the female adult surplus mortality differs among the urban Sct. Mikkel sample and the rural Löddeköpinge and Tirup samples (Boldsen 1984 and 1988). In the Sct. Mikkel sample the surplus mortality is concentrated among the youngest adult females (ages 18-25 years). In the two rural samples a large fraction of the female surplus mortality is due to deaths among more mature women (ages 30-40 years). The patterns of dental and joint disease in the rural Tirup sample indicate that reproductive depletion among women aged 35-45 years was a common associate of death.

        The data presented and analysed in this chapter indicate that adult female surplus mortality is associated with pre-modern agricultural production. It is not possible to determine whether this association is brought about by a heavy work burden, by high fertility or by the microbiological environment of the sedentary, crop-raising communities of Europe from the Neolithic to the late Middle Ages.

        In conclusion, it appears that the steady increase of human adult survival - both male and female - is a fairly new phenomenon in hominid evolution. For the males the crude data analysed here indicate a decline in mortality starting in the Early Middle Ages, during or after the fall of the Roman Empire. This late decline of male mortality is paralleled by the females. The decline is even more pronounced among females than among males. The increase of female mortality leading to the decline of survival from the Mesolithic to the Neolithic is probably a very important shift in the selectional forces shaping our species. If this finding can be substantiated by the inclusion of more European data and information from other parts of the world it might provide guidelines for the analysis and understanding of post-glacial demographic evolution.


Updated by V. Castanova,   March 2000