Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 21.
Published in final edited form as: Archaeol Anthropol Sci. 2026 Mar 18;18(4):s12520-026-02439-z. doi: 10.1007/s12520-026-02439-z

Urban survival, population dynamics and historical transitions in Thessaloniki (316 BC-AD 1500)

Asterios Aidonis 1,2, Christina Kakasa 1,2, Elissavet Ganiatsou 1,2, Panagiota Bantavanou 1,2, Soultana Protopsalti 3, Stavroula Tzevreni 3, Stela Vasileiadou 3, Krino Konstantinidou 3, Christina Papageorgopoulou 1,2,✉
PMCID: PMC7619378  EMSID: EMS217082  PMID: 42626451

Abstract

Cities have historically endured profound challenges while sustaining their populations through adaptive structures and functions. Understanding contemporary urbanization requires situating it within these long-term historical trajectories. This study examines the demographic consequences of historical transitions in Thessaloniki over 1,800 years, based on 846 individuals recovered from peri-urban and intramural cemeteries. Survival analysis was applied to assess mortality risk and life expectancy across six periods: Hellenistic (323–31 BC), Roman (31 BC–AD 324), Early Byzantine (AD 324–842), Middle Byzantine (AD 842–1204), Late Byzantine (AD 1204–1430), and Post-Byzantine (AD 1430–c. 1600). The results demonstrate that from the Hellenistic through the Byzantine periods, mean survival remained largely stable, with no statistically significant fluctuations. In contrast, a pronounced decline in survival was observed during the early Post-Byzantine period, attributable to a succession of catastrophic events between AD 1422 and 1430, including a plague epidemic, an extended siege, and the city’s eventual conquest by the Ottomans. No significant differences in survival were identified between males and females, nor between individuals buried in different cemetery locations, indicating a relatively uniform distribution of urban mortality risks. By reconstructing demographic patterns across an extended temporal framework, this study highlights the interplay of biological and historical forces in shaping urban experience in Thessaloniki across nearly two millennia.

Keywords: Bioarchaeology, Paleodemography, Survival analysis, History of urbanization, Diachronic cities

Introduction

For most of human history, the predominant mode of habitation was settlements with low population density (Klein Goldewijk et al. 2010). Urbanization emerged as a pivotal shift (Algaze 2001; Altaweel 2019; Betsinger and DeWitte 2021; Engel and Brückner 2021; Gates 2011), and within a remarkably brief timespan in terms of evolution, urban living became the predominant form of human settlement. Today, more than half of the world’s population (55%) lives in cities, and this figure is expected to rise to 66% by 2050 (United Nations, 2018), making the study of urban demographics increasingly important in the twenty-first century. However, despite the historical increase in urbanization, past urban environments have been associated with adverse health effects and are often considered detrimental to historical populations (Christ and González Gutiérrez 2022; Crane-Kramer and Buckberry 2020; Newton and Smith 2013; Woods 2003). Empirical research has shown that city dwellers experienced an “urban mortality penalty”, mainly attributed to the prevalence of infectious diseases, periodic food shortages, and lack of adequate hygiene practices (Lewis 2002; Nagaoka et al. 2019; Olga Hernández Espinoza and Márquez Morfín 2015; Roberts and Manchester 2010; Storey 1985; Walter and DeWitte 2017).

While increased urban mortality was a reality, there are also studies indicating a significant degree of variability in the ways these processes affected urban populations (Betsinger and DeWitte 2021). In fact, during their historical transitions, cities have shown remarkable sustainability in providing continuity of habitation. Key elements that ancient cities integrated to achieve sustainability include demographic adaptability, exploitation of geographic perspectives, institution development, and culture (Smith et al. 2021). In demographic terms, it has been suggested that the viability of ancient cities was maintained through increased immigration and higher fertility rates, which countered the losses due to rising mortality rates (Bernard 2016; Bogin 1988; Dyson 1992; Killgrove and Montgomery 2016; Torres et al. 2019).

This pattern reflects a broader principle observed in pre-modern urban societies, where the dynamics of mortality, fertility, and migration were closely intertwined. In early modern European cities, for example, natural decrease was common, yet urban populations persisted or even grew due to substantial in-migration, which replenished the labor force and countered demographic losses (Galley 1995; Sharlin 1978). Similarly, studies of Roman cities show that urban population levels relied heavily on migrants from rural areas and smaller towns, as natural fertility alone could not sustain the cities in the face of high mortality, particularly among children and adolescents (Tacoma 2016).

To better understand these dynamics, the analysis of human skeletal remains provides a unique perspective on the lived experiences of past urban populations, offering critical data on their demographic trends and expanding the historical interpretations. For example, the incorporation of Britain into the Roman Empire after 54–55 BC and the introduction of urbanism, resulted in increased mortality risk for children and older men, thereby challenging the embedded belief concerning the benefits of “Romanization” (Redfern and DeWitte 2011). On the other hand, during the Medieval and Early Modern periods (AD 1100–1850), the age at death of the population of London remained relatively stable around 30 years (Müller-Scheeßel et al. 2024), while after the 17th century, life expectancy and survivorship of adults showed a marked increase, as a result of the second epidemiological transition (DeWitte 2014a; Müller-Scheeßel et al., 2024; Yaussy et al. 2023). Thus, the interplay between historical transitions and urban demography is crucial to understanding the functions that shaped the demographic patterns, and contributed to the resilience of past urban societies.

To this end, the analysis of human remains is a substantial contribution to understanding the demographic shifts that occurred following historical transitions. Current bioarchaeological studies have incorporated survival analysis, a body of statistical procedures that includes both survival functions (e.g., Kaplan–Meier estimates) and hazard models (e.g., Gompertz hazard models and Cox proportional hazards regression) (Clark et al. 2003; Kleinbaum and Klein 2012; Moore 2016), to investigate these demographic trends. For example, Gompertz hazard models have been used to explore differences in mortality risk in the Middle Cumberland Region of Tennessee during the Mississippian period (ca. 1000–1500 AD) (Fojas 2022), while Cox proportional hazards analyses have been applied to investigate the demographic collapse of Indigenous populations in California following European contact (Jones et al. 2021). Additionally, Kaplan-Meier survival analysis has been employed to investigate survivorship in industrial era London(Yaussy et al. 2023), as well as patterns of urbanization in Roman-period Austria (Sammut 2025). These approaches have been proved to be effective in describing the different trends of urbanization in the past (Betsinger et al. 2020; Betsinger and DeWitte 2017; DeWitte 2024; Walter and DeWitte 2017; White et al. 2022). However, rarely have been employed in the context of the circum-Mediterranean region (Biehler-Gomez et al. 2024), where a series of diachronic cities, such as Thessaloniki, Ephesus, Jerusalem, Marseille, or Cordoba, share much in common regarding their urban patterns and historical transitions (Harper 2017).

The present study explores how the synergistic impact of historical phenomena and urbanization affected the population of Thessaloniki. The city of Thessaloniki is a European urban hot spot in the eastern Mediterranean and a representative example of uninterrupted urban development, with a historical course that provides a unique record of transitions, invasions, conquests and epidemic outbreaks. In recent years, large-scale excavations for the metropolitan subway construction unearthed more than 4,000 graves from peri-urban cemeteries of the city (Lambrothanassi et al. 2018). This bioarchaeological archive, dated from the 3rd century BC to the 15th century AD represents the lived history of the population, and provides significant biocultural data for understanding the challenges of urban life.

In this perspective, we evaluated the impact of Thessaloniki’s historical transitions on the population. We studied 664 individuals to reconstruct the city’s demographic trends, in the Hellenistic (323 − 31 BC), Roman (31 BC – AD 324), Early (AD 324–842), Middle (AD 842–1204), Late (AD 1204–1430) and Post Byzantine (AD 1430 - c. 1600) periods, and contextualized the findings with the city’s historical trajectory. As transitional periods with historical complexity and social instability are often related to fluctuation in living conditions, we examined whether these transformations were associated with changes in mortality risk and survival patterns of the population.

Materials and methods

Historical background and cultural context

Thessaloniki was founded by Cassander in 316/5 BC through the amalgamation of 26 surrounding towns and villages. It was named after his wife, Thessaloniki, the daughter of Philip II and half-sister of Alexander the Great. The city became a pivotal economic and political hub due to its strategic position along the trade routes that connected the Aegean Sea with the Balkan hinterland. After 168 BC, under Roman rule, Thessaloniki was designated as the capital of Provincia Macedonia, leading to a significant influx of diverse populations (Antonaras 2019; Nigdelis 2010; Vanderspoel 2010) and fostering urban growth, evidenced by increased residential construction between the 2nd and 4th centuries AD (Salomies 1996; Stefanidou-Tiveriou 2010; Vitti 1996).

Following the establishment of the Byzantine Empire, Thessaloniki retained its political and economic status. The advent of Christianity and construction of numerous churches, many of which have been designated as UNESCO World Heritage Monuments, profoundly influenced both the urban and social topography of the city (Drakoulis 2012). Until the 7th century AD, the city faced repeated raids by the Goths, Huns and Avars, and experienced the outbreak of the Justinian plague (Laiou and Morrisson 2007; Stathakopoulos 2000). The subsequent Slavic invasions led many people to migrate to Thessaloniki and its surroundings in search of refuge (Antonaras 2019). In the next centuries, despite being sacked by Saracens (9th century), engaged in conflicts with the Bulgarians, and undergone a second sack by the Normans in AD 1185, the city remained a military, administrative, and commercial center, exerting strong ideological and cultural influence in the Balkans (Kourkoutidou-Nikolaidou and Tourta 1997; Treadgold 1997). After Constantinople was captured by the Crusaders, leading to political fragmentation of the empire, Thessaloniki became the capital of the short-lived Frankish kingdom (AD 1204–1224). In the second half of the 14th and the early 15th century, recurring plague epidemics, internal strife, and wars caused population decline and eventually the city was occupied by the Turks in 1430 (Laiou and Morrisson 2007).

Summarizing the historical course, four key phases of the city’s development can be distinguished: (a) its establishment during the Hellenistic period, (b) the intensification of urbanization during Roman times, which led to expansion of the inhabited area and the population, (c) the city’s resilience during the turbulent Byzantine era, marked by repeated invasions and epidemics, and (d) the period immediately following its capture by the Ottomans. These historical transitions have been extensively studied from archaeological and historical perspective, but little is known about their impact on the everyday life and demography of the city’s inhabitants.

The skeletal material

Throughout the Hellenistic, Roman, and Early Byzantine periods, the cemeteries of Thessaloniki extended outside the walls, along the roads leading to the main gates (Fig. 1), forming the eastern and western necropoleis of the city (Marki 2006). Around the 7th century AD, cemeteries outside the walls ceased to be used systematically and burials appeared around churches and monasteries within the city walls (Bakirtzis 2003). This shift in burial practices has been attributed to a change in the conceptualization of ad sanctos burials and urban space as well as the insecurity caused by frequent raids (Chatzinikolaou and Terzopoulou 2012; Wataghin 1999). In total, 846 individuals were studied from the east and west peri-urban necropoleis as well as intramural cemeteries of Thessaloniki. However, data from 182 skeletons were excluded from the analysis owing to the absence of secure chronological attribution, resulting in a final sample of 664 individuals. Chronological assignments of skeletons were primarily based on artifact associations and stratigraphic position. Burials that could not be assigned to one of the aforementioned periods were classified as undated. The material under study has been excavated by the Ephorate of Antiquities of the City of Thessaloniki, mainly during the large-scale works for the construction of the metropolitan subway. The skeletal analysis was conducted in accordance with permissions granted by the Ephorate of Antiquities of the City of Thessaloniki, the regional service of the Ministry of Culture responsible for the protection and scientific research of ancient cultural heritage.

Fig. 1.

Fig. 1

Open in a new tab

Map of Thessaloniki depicting the eastern and western cemeteries (shaded red), located outside the city walls (outlined in grey), along with the excavation sites (black dots) representing the origin of the skeletal material. The orange dotted lines represent the roads leading to the main gates of the ancient city, many of which remain in use today (modified from Marki 2006)

Age and sex estimation

Age estimation for non-adults was based on tooth development, diaphyseal length and epiphyseal closure of long bones (Buikstra and Ubelaker 1994; Cunningham et al. 2016; Ferembach et al. 1980). Adult age was estimated based on age-related changes in the pubic symphysis, auricular surface, cranial suture closure and dental wear (Brooks and Suchey 1990; Brothwell 1981; Lovejoy et al. 1985; Meindl and Lovejoy 1985). The skeletons were assigned to one of the following age groups: Fetal (< birth), Infant (0 to 3 year), Child (4 to 12 year), Adolescent (13 to 18 year), Young Adult (19 to 34 year), Middle Adult (35 to 49 year), Old Adult (50 + yr). Individuals lacking precise age markers were grouped in broader categories (nonadult/adult). Sex was estimated using standard methods based on pelvic and cranial morphological variation (Bass 2005; Buikstra and Ubelaker 1994; Ferembach et al. 1980; Phenice 1969). The skeletons were initially categorized as female, probably female, indeterminate, probably male, or male. However, following standard bioarchaeological practice, individuals were grouped into three categories for statistical analysis: male, female, and indeterminate (which included nonadults, individuals with non-diagnostic features, or those with poor skeletal preservation).

Demographic analysis

A comparison of survivorship patterns over time was conducted based on age-at-death estimates derived from skeletal remains. All the analyses were run in R and SPSS. The mortAAR package (Mueller-Scheessel et al. 2025) was used to evaluate the representativeness of the nonadults in the dataset. The relevant function of the mortAAR package uses indices to compare the proportions observed among nonadult age groups with those documented in contemporary, comparable populations (Bocquet-Appel and Masset 1977; Taylor and Oxenham 2025; Weiss and Wobst 1973). The impact of bone preservation on the representativeness of the sample was investigated by calculating the preservation index following the method proposed by Stojanowski et al. (2002). Additionally, the representativeness of the sexes within the dataset was ascertained by calculating the sex ratio (SR=100∗malesfemales) of the skeletal sample.

We then constructed life tables using the relevant function of the mortAAR package, which employs a statistical approach that assumes a uniform probability of individuals being distributed across the five-year age classes of the initial broader age-at-death estimations. An individual aged within the age range of 20 to 34 years was distributed by adding 1/3 to each of the three adjusted age classes. In cases of broader age estimation, smaller fractions were added to the relevant age classes (Acsadi and Nemeskeri 1970). For old adult individuals, whose maximal age is in fact open-ended, we assumed an upper limit of 80 years.

At this step we excluded the nonadults aged 0–9 years to address the observed underrepresentation of infants and children commonly present in archaeological skeletal populations (Jones et al. 2021; Kelmelis and DeWitte 2021), as well as the adults of unknown age and sex, to avoid the potential bias in the results. We then used the median age of each five-year class above 10, repeated according to the number of deaths in each class, to build the age distributions for comparing age-at-death patterns and conducting the survival analysis.

Mann–Whitney U tests were first employed to assess whether significant differences existed in age-at-death distributions between periods (Chamberlain 2006). The survival package (Therneau 2024; Therneau and Grambsch 2000) was then used to perform a Cox proportional hazards analysis in order to evaluate variations in mortality risk. We used the defined historical periods, the sex and the cemetery as covariates that may influence survival. The proportional hazards assumption for each covariate used in the model was checked with the relevant function of the survival package. Finally, we conduct a series of Kaplan–Meier analyses (Kaplan and Meier 1958) with the associated log-rank tests to investigate survivorship across and within the time periods under study and between females and males. Assuming that the eastern cemetery might represent higher status burials due to its proximity to the Galerius Palace Complex, we also examined survivorship between the burials in the city’s eastern and western cemetery.

It has been discussed by previous studies that age-at-death distributions in ancient cemeteries may better reflect fertility than mortality. Research has demonstrated that there is a strong correlation between the proportion of juveniles in cemetery samples and fertility of the once living population (Bocquet-Appel 2002; Bocquet-Appel & Naji, 2006; Buikstra et al. 1986; McFadden and Oxenham 2018; Paine 1989; Sattenspiel and Harpending 1983). Actually, an increase in fertility results in a higher proportion of nonadults and a lower mean age at death. Conversely, a decrease in fertility leads to a higher proportion of adults and a higher mean age at death (Betsinger et al. 2020; Betsinger and DeWitte 2017; Jones et al. 2021; Zoeller et al. 2022). A number of fertility proxies have been developed (Bocquet-Appel, 2002; Buikstra et al. 1986; McFadden and Oxenham 2018), and have been used to evaluate if the observed changes in survival actually represent changes in fertility (Walter and DeWitte 2017; Zoeller et al. 2022). To evaluate the effect of fertility on survivorship in the chronological and cultural periods under study, we employed the P indicator (D5−19//D5+ ratio), which is positively correlated to birth and growth rates (Bocquet-Appel, 2002; Bocquet-Appel & Naji, 2006). In order to evaluate the significance of diachronic changes between the periods, we used the 95% comparison intervals as proposed by Buikstra (Buikstra et al. 1986).

CI=D30+D5+±m05,k∗,∞.125D5+

where m is read from a table of the studentized maximum modulus, at a two-tailed alpha level of 0.05, with k* as the (N(N− 1)/2 for N samples, and under infinite (∞) degrees of freedom (Stoline and Ury 1979). Furthermore, it has been suggested that when the population growth rate (r) is low, the P indicator provides a direct measure of mortality (Barbiera et al. 2018). Accordingly, for growth rates deviating from ± 3‰, we adjusted the P values to account for shifts in the population’s age structure resulting from changes in birth rates, using the d* indicator as described by Barbiera et al. (2018).

Results

Nonadults under the age of ten were found to be under-represented in the Hellenistic, Early Byzantine and Post Byzantine data sets, while the Roman, Middle and Late Byzantine samples could be characterized representative, by satisfying one of the conditions either of Weiss and Wobst (1973) or Bocquet-Appel and Masset (1977). However, none of the subsamples meet the criterion of Taylor and Oxenham (2025) (see Table S1). As for bone preservation, age was found to affect it significantly (p < 0.001); infants aged 0–3 years (IS = 0.0831) and children aged 3–12 years (IS = 0.0883) present the lowest bone preservation, with skeletal completeness of 8.31% and 8.83% respectively (Table S2-S3). Owing to this underrepresentation, the 0–9 age classes were excluded from both the comparison of age-at-death distributions and the survival analysis.

The age-at-death distributions for each chronological period are shown in Table 1, while the age at death distributions in 5-year age classes are presented in the form of life tables in Table S4. Life expectancy at age twenty (e20) declines from 24.4 years in the Hellenistic period to 21.86 years in the Roman period. It remains relatively stable from the Early Byzantine (20.34 years) to the Late Byzantine period (20.85 years), before dropping sharply in the Post Byzantine period to its lowest value of 12.65 years. The comparison of age-at-death distributions using the Mann-Whitney U test revealed that there were no significant differences between the Hellenistic, Roman and all three Byzantine groups. However, the Post Byzantine group did show a significant differentiation (p = 0.001) as depicted in Fig. 2 and Table S5. The examination of the sex ratio by chronological period indicates that females outnumber males in the Hellenistic (SR = 70.0), Roman (SR = 88.9), Early (SR = 68.6) and Middle Byzantine (SR = 52.6) periods. Conversely, there is a skew towards males in the Late Byzantine (SR = 136.4) and even more in the Post Byzantine period (SR = 241.7), where the underrepresentation of women is significant (Table S6; p = 0.012).

Table 1. Distribution of individuals from Thessaloniki per age group and chronological period.

Hellenistic Roman Early Byzantine Middle Byzantine Late Byzantine Post-Byzantine Total
(323 − 31 BC) (31 BC − AD 324) (AD 324−842) (AD 842−1204) (AD 1204−1430) (AD 1430 − c. 1600)
N % N % N % N % N % N % N %
Fetal (<birth) 0 0,0% 1 0,4% 0 0,0% 0 0,0% 0 0,0% 0 0,0% 1 0,2%
Infant (0–3 years) 2 3,7% 29 12,3% 1 1,3% 4 12,5% 4 5,8% 2 3,3% 42 7,9%
Child (4–12 years) 5 9,3% 56 23,8% 18 22,5% 3 9,4% 7 10,1% 11 18,0% 100 18,8%
Adolescent (13–18 years) 6 11,1% 16 6,8% 5 6,3% 2 6,3% 6 8,7% 8 13,1% 43 8,1%
Young adult (19–34 years) 16 29,6% 61 26,0% 30 37,5% 16 50,0% 27 39,1% 25 41,0% 175 33,0%
Middle adult (35–49 vears) 12 22,2% 53 22,6% 22 27,5% 6 18,8% 19 27,5% 15 21.6% 127 23,9%
Old Adult (50 + years) 13 24,1% 19 8,1% 4 5,0% 1 3,1% 6 8,7% 0 0,0% 43 8,1%
Total 54 100,0% 235 100,0% 80 100,0% 32 100,0% 69 100,0% 61 100,0% 531 100,0%
Nonadult (0–19) 0 0,0% 1 1,7% 1 3,7% 0 0,0% 0 0,0% 0 0,0% 2 1,5%
Adult (20 +years) 9 100,0% 57 98,3% 26 96,3% 12 100,0% 25 100,0% 2 100,0% 131 98.5%
Total 9 100,0% 58 100,0% 27 100,0% 12 100,0% 25 100,0% 2 100,0% 133 100,0%
Total sample 63 293 107 44 94 63 664
Open in a new tab

Fig. 2. Visual representation of the Mann-Whitney U test results, illustrating age-at-death distributions across groups from the Hellenistic to the Post-Byzantine period.

Fig. 2

Open in a new tab

The results of the Cox proportional hazards model indicate that the covariate of period is significantly related to elevated risk of death (p < 0.001), while sex and burial topography are not (Table 2). Overall, the p-values for all three tests (likelihood, Wald, and score) indicate that the model is significant. When testing the proportionality assumption, there is no evidence to suggest that the proportional hazards assumption is violated, either for individual variables or globally, as all the covariates are not significant (Table S7).

Table 2. Results of Cox proportional hazard model for the covariates of sex, chronological period and topography of the burial (east and West cemetery).

Hazard Ratio SE 95% CI z p-value
Period 1.210 0.041 1.115–1.313 4.586 < 0.001
Cemetery 0.879 0.084 0.745–1.037 –1.530 0.126
Sex 1.076 0.109 0.869–1.334 0.677 0.498
Concordance = 0.575 (SE = 0.02)
Likelihood ratio test = 21.57 on 3 df, p < 0.001
Wald test = 22.63 on 3 df, p < 0.001
Score (logrank) test=22.8 on 3 df, p < 0.001
Open in a new tab

The results of the Kaplan-Meier analyses revealed difference in survivorship over time (Table 3, p < 0.001). Pairwise comparisons of the chronological samples indicate broadly comparable survival times from the Hellenistic through the Late Byzantine periods, suggesting demographic stability across these centuries. In contrast, the sharp decline documented in the Post Byzantine period is statistically significant, marking a clear disruption in population survival patterns (Fig. 3). This is also visually evident in the survival curves, which display reduced survivorship among Post Byzantine individuals (Fig. 4).

Table 3. The results of the Kaplan-Meier survival analysis by time period, with the mean and median survival times, their corresponding 95% confidence intervals and the results of the associated log-rank test.

N Mean survival time (years) SE Median survival time (years) Median survival time 95% CI
Hellenistic 57 39.623 2.359 37.500 32.500–47.500
Roman 208 39.495 1.193 37.500 32.500–42.500
Early Byzantine 91 38.269 1.739 37.500 32.500–42.500
Middle Byzantine 37 37.229 2.394 32.500 27.500–42.500
Late Byzantine 84 38.333 1.851 37.500 32.500–42.500
Post Byzantine 51 28.088 1.478 27.500 22.500–32.500
Log-rank test: Chisq: 28.4
df: 5
p < 0.0001
Open in a new tab

Fig. 3.

Fig. 3

Open in a new tab

Heatmap plot showing the p-values from the pairwise Kaplan Meier comparisons between the chronological groups. Color intensity reflects the p-value, with lighter shades indicating significant differences (p < 0.05)

Fig. 4. Kaplan–Meier survival curves with 95% confidence intervals indicate a significantly lower survivorship in the Post Byzantine sample.

Fig. 4

Open in a new tab

A second series of Kaplan-Meier analyses was performed to examine survival patterns by sex across the chronological periods. The results indicate that male and female mean survival times were broadly comparable in all periods, with no statistically significant differences (Fig. 5 and Table S8). The survival times between the eastern, western, and intramural cemeteries by period, also show no significant differences (Fig. 6 and Table S9). The P indicator (D5-19/D5+) shows fluctuations across chronological periods, starting at 0.1379 in the Hellenistic period, rising to 0.2017 in the Roman and 0.2039 in the Early Byzantine periods, then declining to 0.1250 in both the Middle and Late Byzantine periods. In the Post Byzantine period, the ratio increased markedly to 0.2632. However, the overlapping 95% comparison intervals suggest that these differences are not statistically significant. The application of the d* indicator following Barbiera et al. (2018), further reduces these variations, resulting in a smoother demographic profile across the examined periods. Across chronological periods, the r* indicator measured 0.1604 in the Hellenistic, 0.1792 in the Roman, 0.1814 in the Early Byzantine, 0.1565 in both the Middle and Late Byzantine, and 0.2092 in the Post Byzantine period. (Fig. 7; Table S10).

Fig. 5. Kaplan–Meier survival curves illustrating female and male survivorship across chronological periods in Thessaloniki.

Fig. 5

Open in a new tab

Fig. 6. Kaplan–Meier survival curves illustrating survivorship across east, west, and intramural cemeteries by time period in Thessaloniki.

Fig. 6

Open in a new tab

Fig. 7. The P indicator (blue line) with 95% comparison intervals (light blue shaded area) and the r* indicator (red line) by chronological period in Thessaloniki.

Fig. 7

Open in a new tab

Discussion

The living experience of city dwellers is shaped by the cumulative effects of historical transitions on urban infra-structure and the social fabric. The present study employed a paleodemographic analysis to examine the impact of historical processes on the demographic profile of Thessaloniki during the Hellenistic, Roman, Byzantine and Post Byzantine periods.

The analysis of bone preservation in the sample of Thessaloniki revealed that poor preservation may have affected representation. This phenomenon was observed in all chronological periods except the Roman. However, this is a common finding in bioarchaeological research and is usually attributed to: (a) cultural beliefs influencing burial practices, (b) differential preservation of the fragile infant bones, and (c) limited archaeological research (Saunders and Barrans 1999). Furthermore, children were frequently excluded from formal burial sites during Greek antiquity (Dimakis 2021; Lagia 2007) due to their late status recognition (Lagia 2007; Lancy 2014). Atypical funerary treatment of infants can be traced back to the Mycenaean period (Lebegyev 2009) and has been detected in Mendi, Halkidiki (Vokotopoulou 1989), the Agora of Athens in the so-called “well of the babies” (Liston and Rotroff 2013), and in Kylindra, Astypalaia (Hillson 2009). Although a comparable funerary treatment has yet to be identified in Thessaloniki, it cannot be dismissed. It can therefore be posited that the underrepresentation of infants and children in the Hellenistic, Byzantine and Post Byzantine sample from Thessaloniki may be influenced by cultural and taphonomic determinants, as well as the limited nature of the excavations.

The sex ratios observed in Thessaloniki vary across the different periods, but it is only in the Post Byzantine period women are significantly underrepresented. Most bioarchaeological studies indicate a systematic bias favoring males, which can be attributed to burial practices, biased methods of sex estimation, and the better preservation of male skeletons (Bennike 1985; Biehler-Gomez et al. 2022; Cintas-Peña and Herrero-Corral 2020; Weiss 1972). However, in demographic terms, the sex ratio is influenced by births, deaths, and migration (Hesketh and Xing 2006). Urban migration is often considered primarily labor migration, which depends on the availability of employment opportunities. For example, in Rome, most jobs were in manual labor and predominantly occupied by male workers (Erdkamp 2008), although alternative perspectives are also being explored (Prowse 2016). In Post Byzantine Thessaloniki, the marked underrepresentation of females is hardly explained solely by the demands of the city’s labor market. It is more reasonable to interpret this phenomenon within the broader context of the demographic upheavals between 1422 and 1430, a period marked by a plague outbreak, prolonged siege conditions, and ultimately the Ottoman conquest. The depopulation of Thessaloniki during the Ottoman siege, followed by the forced resettlement of new populations after the city’s conquest, may triggered substantial demographic restructuring. This process likely produced a male-biased population structure, which shaped both by the predominance of men among the city’s defenders and by the composition of incoming groups, many of whom were recruited for military service and other labor-intensive activities traditionally dominated by males.

The results of the Cox proportional hazards analysis indicate that chronological period is a significant predictor of mortality risk, highlighting temporal shifts in population survival. Pairwise Kaplan–Meier comparisons show no statistically significant differences in survival times between the Hellenistic, Roman, and all Byzantine periods, suggesting relative stability in mortality patterns over this extended timeframe. In contrast, the Post Byzantine sample exhibits a marked and statistically significant decline in survival. The P indicator (D5−19//D5+) reflects temporal variation, with higher values in the Roman, Early Byzantine, and particularly the Post Byzantine period. Nevertheless, overlapping comparison intervals suggest that these differences are not statistically significant and thus do not affect survival. Beyond reflecting birth and growth rates (Bocquet-Appel 2002; Bocquet-Appel & Naji, 2006), in the case of stable populations (r = ± 3‰), this measure can serve as a direct proxy for mortality (Barbiera et al. 2018). However, the assumption of population stability does not hold during periods of mortality crises, as deaths could rise far above normal levels and reshape demographic dynamics (Barbiera et al. 2018). In Thessaloniki, the growth rate calculated using the P indicator (Bocquet-Appel, 2002) fluctuates above the ± 3‰ threshold (Table S10). When adjusted using the d* indicator (Barbiera et al. 2018), the mortality pattern mirrors the overall trend of the P indicator with higher values also in the Roman, Early Byzantine, and Post Byzantine period, but shows a much smoother distribution across periods, with less pronounced fluctuations.

An examination of life expectancy at birth (e0) reveals fluctuations between the periods under study (Table S4). It is important to note that life expectancy at birth (e0) in archaeological samples is strongly influenced by the systematic underrepresentation of children under five years of age, either due to differential burial practices or poor preservation of fragile remains (Chamberlain 2006; Hoppa, 2002; McFadden 2021), Indeed, such underrepresentation is evident in our skeletal sample as well. While e0 values are reported, they should be interpreted with caution as they primarily reflect sampling and preservation biases rather than actual population demographic conditions. For this reason, we focus our interpretation on life expectancy at age 20 (e20), which provides a more robust indicator less affected by these biases. Life expectancy at age 20 (e20) declines from the Hellenistic to the Roman period, remains relatively stable during the Byzantine period, and reaches its lowest point in the Post Byzantine period, following a pattern similar to that observed in the Kaplan Meier analysis and the P indicator.

The comparative analysis of the above demographic data along with the historical transitions that shaped Thessaloniki’s urban trajectory yields interesting insights. Archaeological sources demonstrate that during the Roman period, the economic growth and the upgrade of the city to a provincial capital resulted in a substantial population growth (Antonaras 2019; Nigdelis 2010; Vanderspoel 2010). On the basis of inhabited area, the population at this time is estimated to 70,000 inhabitants (Hanson et al. 2019), whereas the city continued to be a crowded center during the Byzantine period. Although, urban intensification has been a contributing factor to increased mortality rates, largely attributable to elevated exposure to environmental pollutants and the dissemination of infectious pathogens (Capasso 2000; DeWitte 2021; Havlíček and Morcinek 2016; Ledger et al. 2018; Mackenbach 2021; Mitchell 2017; Novak 2010; Reyes et al. 2013), our results do not support a significant correlation between the periods of intense urbanization and mortality. One possible reason for the survivorship patterns observed during these periods is the implementation of policies and legal measures aimed at improving sanitation and reducing pollution, which may have lessened some of the health risks associated with dense urban living. Archaeological and historical sources support that fundamental sanitation infrastructure (Demetriou 2002; Kaminiates 2000; Karidas 1999), such as baths (Arvanitidou 2018), and waste management mechanisms (Bakirtzis 2007; Kaiafa-Saropoulou and Karadedos 2011; Karpozilos 1989) were constructed. The residents of Thessaloniki were required to maintain the cleanliness of areas outside private property (Marki 2004), and groups of workers were responsible for supervising streets and pavements (Bakirtzis 2007). Such practices would have functioned as regulatory mechanisms for environmental pressures resulting from the negative effects of increasing population density and contributed to the resilience of the city.

Throughout the Byzantine period, the city endured the impact of the Justinianic plague, repeated sieges, looting and wars involving Saracens, Normans, Franks, and Bulgarians, as well as episodes of internal unrest such as the Zealot movement. While such crises would have inevitably impacted the population, our results show no clear evidence of survival declines during these times. Detecting the demographic impact of specific crisis events is inherently challenging, as urban populations often recovered rapidly (DeWitte 2014b; Galanaud et al. 2020; Galley 1995), and such events are more effectively investigated in catastrophic cemeteries rather than in attritional burial grounds. In Thessaloniki, despite repeated crises and episodes of expected heightened mortality, the city’s population appears to have rebounded, underscoring the resilience of urban structures. Consequently, mortality from these crises, may have caused temporary disruptions not detected in attritional cemeteries, with long-term demographic patterns largely determined by enduring urban dynamics rather than short-term shocks.

In contrast, the Post Byzantine dataset shows a marked decline in mean survival time, along with a pronounced absence of individuals over 50 years of age and a highly skewed sex ratio favoring males. This pattern may reflect the cumulative impact of a series of historical events. The Venetian occupation of Thessaloniki (AD 1423–1430), the plague outbreak of 1421–1422, the prolonged Ottoman siege after 1422, and the city’s final capture by the Ottomans in 1430 all contributed to a gradual demographic decline (Laiou and Morrisson 2007; Matschke 2002; Tsiamis et al. 2011; Vacalopoulos 1973). Following these events, Thessaloniki was nearly abandoned, with historical sources estimating the population at only around 7,000 inhabitants (Reba et al. 2016). Two years later, Ottoman authorities initiated repopulation efforts, relocating Muslims and Greeks from surrounding villages (Vacalopoulos 1973). Population displacement in the newly conquered territories was a common practice for the Ottoman authorities in order to secure the new lands and to restart the economic activity (Hooper 2003; Lowry 1986). Therefore, it is reasonable to assume that the Post Byzantine skeletal sample reflects the impact of these crises on the city’s population structure, with the significantly lower mean survival age of an immigrant population predominantly composed of young and middle adult males.

The analysis of survival across the city’s subpopulations revealed no significant differences, with males and females exhibiting comparable mean survival times across periods. In preindustrial societies, females generally exhibited lower survival rates than males, largely due to complications from pregnancy, childbirth and greater exposure to pathogens from spending more time indoors (Boldsen and Paine 1995; Cheng and Nelson 2018). By contrast, in most contemporary populations, women outlive men, a reversal known as the sex morbidity–mortality paradox. This pattern emerged in Europe only in the past few centuries, with female life expectancy first surpassing that of males in 1751 in Sweden, 1841 in England and Wales, and 1889 in Italy (Barford et al. 2006; Zarulli et al. 2018). This male excess in mortality has been attributed to the long-term demographic and epidemiological transition, with the reduction of infections and the increase of chronic diseases (Beltrán-Sánchez et al. 2015).

The bioarchaeological evidence on sex differences in mortality shows considerable variability. For instance, in Milan, females exhibited significantly lower mean survival times in Roman times, whereas no differences were observed in the Early and Late Middle Ages (Biehler-Gomez et al. 2022). Similarly, in Medieval Ireland, mortality risks did not differ between sexes (Ham 2025). In contrast, female survival rates exceeded those of males in Roman Dorset (Redfern et al. 2015). This heterogeneity likely reflects context-specific factors, which do not always align with the commonly assumed pattern of higher female mortality in preindustrial societies. The comparable mortality profiles of males and females in Thessaloniki likely reflect mortality risks that were broadly shared across the population irrespective of sex.

Additionally, the evaluation of the mean survival time between the eastern, western and the intramural cemeteries of Thessaloniki revealed also no significant differences. Previous bioarchaeological studies have demonstrated that the residents of a city may experience urban living in diverse ways and the potential benefits or disadvantages of urbanization may not distributed evenly in the population (Betsinger and DeWitte 2021; Crane-Kramer and Buckberry 2023; White et al. 2022). However, such patterns were not observed in our sample, as the current data indicate a uniform distribution of mortality risks across demographic groups in each chronological period. However, further investigation is needed, as these results may have been influenced by the limitations of the dataset, particularly the fragmentary nature of the excavations, which were often confined to restricted areas within the extensive cemeteries of Thessaloniki, and may therefore have obscured more subtle differences.

Conclusion

This study presents a diachronic analysis of survivorship in the population of Thessaloniki, and provides informative results on the demographic impact the city experienced during its historical transitions. The findings suggest that from the Hellenistic period to the end of the Byzantine era, there were no significant changes in the average survival time. Despite the occurrence of historically documented periods of crisis, mainly during the Byzantine period, this is not reflected in the demographic patterns of the population. However, there was a statistically significant decline detected in the Post Byzantine period, showing that the city’s demographic composition changed significantly only when the population experienced the consequences of catastrophic events between 1422 and 1430. No significant difference in survival was detected between the sexes and between the individuals of the city’s cemeteries, indicating a uniform distribution of urban risks in each chronological period. Despite its limitations as every bioarchaeological study, this research contributes to a growing body of knowledge and provides a valuable record of the diachronic population dynamics of Thessaloniki by presenting a demographic chart of the city over a period of 1,800 years.

Supplementary Material

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s12520-026-02439-z.

Supplementary Information

Acknowledgements

We would like to thank the Ephorate of Antiquities of the City of Thessaloniki, for the excellent cooperation and permission to study the skeletal material.

Funding

Open access funding provided by HEAL-Link Greece. A.A., E.G., P.B. and C.K. were funded by CityLife, an ERC Consolidator Grant project awarded to C.P., and has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (ERC, CITYLIFE, 101126337). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. This research was also co-financed by the European Regional Development Fund of the European Union and Greek national funds through the Operational Program Competitiveness, Entrepreneurship and Innovation, under the call RESEARCH – CREATE – INNOVATE (project title: ECHOES-Development of a methodology for the digital reconstruction of ancient human biographies through the study of archaeo-anthropological material, project code: T2EDK-00152). A.A. was awarded a State Scholarships Foundation (IKY) grant by Greece and the European Social Fund through the Operational Programme: Human Resources Development, Education and Lifelong Learning, Act: “Enhancing Human Resources Research Potential by undertaking a Doctoral Research, subaction 2”, implemented by the State Scholarships Foundation (IKY) (MIS 5113934).

Declarations

Author contributions A.A. and C.K. studied the skeletal material, A.A. compiled reference datasets, performed statistical analyses, prepared figures and tables and wrote the original draft. E.G., P.B., C.K., contributed to contextualization and study of skeletal material. S.P., S.T., S.V., K.K., provided material, and contributed to contextualization of the data. C.P. designed and supervised the project, provided resources and funding, contributed to inference of the data and edited the original draft. All authors reviewed the final manuscript.

Ethical approval The study of the archaeological human remains is the subject of a memorandum of cooperation between the Ephorate of Antiquities of the City of Thessaloniki and the Laboratory of Biological Anthropology of the Democritus University of Thrace. It was conducted in accordance with the ethical and scientific principles that govern the study of archaeological human remains.

Competing interests The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to juris-dictional claims in published maps and institutional affiliations.

Data availability

This article and the SI include all data generated and analyzed during this study.

References

  1. Acsadi G, Nemeskeri I. History of Human Life Span and Mortality. Akademiai Kiado; 1970. [Google Scholar]
  2. Algaze G. Initial social complexity in Southwestern Asia: the Mesopotamian advantage. Curr Anthropol. 2001;42(2):199–233. doi: 10.2307/3596412. [DOI] [Google Scholar]
  3. Altaweel M. Southern Mesopotamia: water and the rise of urbanism. WIREs Water. 2019 doi: 10.1002/wat2.1362. [DOI] [Google Scholar]
  4. Antonaras AC. Arts, crafts and trades in Ancient and Byzantine Thessaloniki archaeological, literary and epigraphic evidence. Mainz Verlag des Römisch-Germanischen; 2019. [DOI] [Google Scholar]
  5. Arvanitidou I. Σέµογλου A, Aρβανιτίδου IΠ, Γούναρη EΓ, editors. Public baths in the early Byzantine Thessaloniki. Λεπέτυµνος Mελέτες Aρχαιολογίας και Tέχνης στη µνήµη του Γεωργίου Γούναρη. 2018:25–39. Σταµούλη. [Google Scholar]
  6. Bakirtzis C. The urban continuity and size of late Byzantine Thessalonike. Dumbarton Oaks Pap. 2003;57:35–64. doi: 10.2307/1291875. [DOI] [Google Scholar]
  7. Bakirtzis C. Byzantium, Pliska, and the Balkans. De Gruyter; 2007. Imports, exports and autarky in Byzantine Thessa-lonike from the seventh to the tenth century; pp. 89–118. [DOI] [Google Scholar]
  8. Barbiera I, Castiglioni M, Dalla Zuanna G. A synthetic measure of mortality using skeletal data from ancient cemeteries: the d index. Demogr Res. 2018;38:2053–2072. doi: 10.4054/DemRes.2018.38.65. [DOI] [Google Scholar]
  9. Barford A, Dorling D, Smith GD, Shaw M. Life expectancy: women now on top everywhere. BMJ. 2006;332(7545):808. doi: 10.1136/bmj.332.7545.808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bass WM. In: Human Osteology: A Laboratory and Field Manual. 5th edn. Bass WM, editor. Missouri Archaeological Society; 2005. [Google Scholar]
  11. Beltrán-Sánchez H, Finch CE, Crimmins EM. Twentieth century surge of excess adult male mortality. Proc Natl Acad Sci U S A. 2015;112(29):8993–8998. doi: 10.1073/pnas.1421942112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bennike P. Paleopathology of Danish skeletons: A comparative study of demography, disease and injury. Akademisk Forlag; 1985. [Google Scholar]
  13. Bernard SG. Migration and Mobility in the Early Roman Empire. BRILL; 2016. Food Distributions and Immigration in Imperial Rome; pp. 50–71. [DOI] [Google Scholar]
  14. Betsinger TK, DeWitte S. Trends in mortality and biological stress in a medieval Polish urban population. Int J Paleopathol. 2017;19:24–36. doi: 10.1016/j.ijpp.2017.08.008. [DOI] [PubMed] [Google Scholar]
  15. Betsinger TK, DeWitte SN. Toward a bioarchaeology of urbanization: demography, health, and behavior in cities in the past. Am J Phys Anthropol. 2021;175(S72):79–118. doi: 10.1002/ajpa.24249. [DOI] [PubMed] [Google Scholar]
  16. Betsinger TK, DeWitte SN, Justus HM, Agnew AM. In: The bioarchaeology of urbanization: The biological, demographic, and social consequences of living in cities. Betsinger TK, DeWitte SN, editors. Springer; 2020. Frailty, Survivorship, and Stress in Medieval Poland: A Comparison of Urban and Rural Populations; pp. 223–243. [DOI] [Google Scholar]
  17. Biehler-Gomez L, Mattia M, Mondellini M, Palazzolo L, Cattaneo C. Differential skeletal preservation between sexes: a diachronic study in Milan over 2000 years. Archaeol Anthropol Sci. 2022;14(8):147. doi: 10.1007/s12520-022-01616-0. [DOI] [Google Scholar]
  18. Biehler-Gomez L, Yaussy S, Moro C, Morandini P, Mondellini M, Petrosino D, Mattia M, Del Bo B, Cattaneo C. Unveiling the female experience through adult mortality and survivorship in Milan over the last 2000 years. Sci Rep. 2024;14(1):21328. doi: 10.1038/s41598-024-71607-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Bocquet-Appel J. Paleoanthropological traces of a neolithic demographic transition. Curr Anthropol. 2002;43(4):637–650. doi: 10.1086/342429. [DOI] [Google Scholar]
  20. Bocquet-Appel J, Masset C. Estimateurs En paléodémographie. L’Homme. 1977;19(4):65–90. [PubMed] [Google Scholar]
  21. Bocquet-Appel J, Naji S. Testing the hypothesis of a worldwide neolithic demographic transition. Curr Anthropol. 2006;47(2):341–365. doi: 10.1086/498948. [DOI] [Google Scholar]
  22. Bogin B. In: Biological Aspects of Human Migration. Mascie-Taylor C, Lasker G, editors. Cambridge University Press; 1988. Biological Aspects of Human Migration; pp. 90–127. [Google Scholar]
  23. Boldsen JL, Paine RR. In: Exceptional Longevity: From Prehistory to the Present. Jeune B, Vaupel JW, editors. Odense University; 1995. The evolution of human longevity from the Mesolithic to the Middle Ages: An analysis based on skeletal data; pp. 25–36. [Google Scholar]
  24. Brooks S, Suchey JM. Skeletal age determination based on the Os pubis: a comparison of the Acsádi-Nemeskéri and Suchey-Brooks methods. Hum Evol. 1990;5(3):227–238. doi: 10.1007/BF02437238. [DOI] [Google Scholar]
  25. Brothwell RD. Digging Up Bones: The Excavation, Treatment, and Study of Human Skeletal Remains. Cornell University Press; 1981. [Google Scholar]
  26. Buikstra JE, Ubelaker DH. (Arkansas Archeological Survey research series).Standards for Data Collection from Human Skeletal Remains. 1994;44(44) [Google Scholar]
  27. Buikstra JE, Konigsberg LW, Bullington J. Fertility and the development of agriculture in the prehistoric Midwest. Am Antiq. 1986;51(3):528–546. doi: 10.2307/281750. [DOI] [Google Scholar]
  28. Capasso L. Indoor pollution and respiratory diseases in ancient Rome. Lancet. 2000;356(9243):1774. doi: 10.1016/S0140-6736(05)71971-1. [DOI] [PubMed] [Google Scholar]
  29. Chamberlain A. Demography in Archaeology. Cambridge University Press; 2006. [Google Scholar]
  30. Chatzinikolaou K, Terzopoulou D. Adam-Veleni P, Terzopoulou D, editors. The cemeteries of Roman Thessaloniki. Field, House, Garden, Grave Archaeological Museum of Thessaloniki. 2012:105–122. [Google Scholar]
  31. Cheng CJ, Nelson JF. Physiological basis for sex-specific differences in longevity. Curr Opin Physiol. 2018;6:57–64. doi: 10.1016/j.cophys.2018.04.003. [DOI] [Google Scholar]
  32. Christ M, González Gutiérrez C. Introduction: death and the city in premodern Europe. Mortality. 2022;27(2):129–143. doi: 10.1080/13576275.2022.2063529. [DOI] [Google Scholar]
  33. Cintas-Peña M, Herrero-Corral AM. Missing prehistoric women? Sex ratio as an indicator for analyzing the population of Iberia from the 8th to the 3rd millennia B.C. Archaeol Anthropol Sci. 2020;12(11):263. doi: 10.1007/s12520-020-01215-x. [DOI] [Google Scholar]
  34. Clark TG, Bradburn MJ, Love SB, Altman DG. Survival analysis part I: basic concepts and first analyses. Br J Cancer. 2003;89(2):232–238. doi: 10.1038/sj.bjc.6601118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Crane-Kramer G, Buckberry J. In: The bioarchaeology of urbanization: The biological, demographic, and social consequences of living in cities. Betsinger TK, DeWitte SN, editors. Springer; 2020. Is the Pen Mightier than the Sword? Exploring Urban and Rural Health in Victorian England and Wales Using the Registrar General Reports; pp. 403–433. [DOI] [Google Scholar]
  36. Crane-Kramer G, Buckberry J. Changes in health with the rise of industry. Int J Paleopathol. 2023;40:99–102. doi: 10.1016/j.ijpp.2022.12.005. [DOI] [PubMed] [Google Scholar]
  37. Cunningham C, Scheuer L, Black S. Developmental Juvenile Osteology. Second. Second Edition; 2016. Developmental Juvenile Osteology. [Google Scholar]
  38. Demetriou O. Aποχωρητήρια και σχετικά θέµατα υγιεινής στην ελληνική κατοικία της αρχαίας, βυζαντινής και νεώτερης εποχής [Aριστοτέλειο Πανεπιστήµιο Θεσσαλονίκης] 2002 doi: 10.26262/heal.auth.ir.2210. [DOI] [Google Scholar]
  39. DeWitte SN. In: Modern Environments and Human Health. 1st ed. Zuckerman MK, editor. Wiley; 2014a. Modeling the Second Epidemiologic Transition in London: Patterns of Mortality and Frailty During Industrialization; pp. 35–53. [DOI] [Google Scholar]
  40. DeWitte SN. Mortality risk and survival in the aftermath of the medieval Black Death. PLoS ONE. 2014b;9(5) doi: 10.1371/journal.pone.0096513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Dewitte SN. In: The Routledge Handbook of the Bioarchaeology of Climate and Environmental Change. Schug GR, editor. Routledge; 2021. Urban environments: Demography, epidemiology, and the role of climate change in determining health outcomes. [DOI] [Google Scholar]
  42. DeWitte SN. Medieval monastic health: Variation in skeletal signs of inflammation and developmental stress between religious orders in London. Am J Hum Biology. 2024:e24052. doi: 10.1002/ajhb.24052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Dimakis N. In: Children in Antiquity Perspectives and experiences of childhood in the ancient mediterranean. Beaumont LA, Dillon M, Harrington N, editors. Routledge; 2021. Child death in the Hellenistic world; pp. 506–518. [Google Scholar]
  44. Drakoulis D. Historical-Geographical references and cartographic semiosis of Thessaloniki during the late antiquity period (in Greek) Byzantiaka. 2012;30:11–43. [Google Scholar]
  45. Dyson SL. Community and society in Roman Italy. Johns Hopkins University; 1992. [Google Scholar]
  46. Engel M, Brückner H. Middle East and North Africa. BRILL; 2021. Holocene Climate Variability of Mesopotamia and its Impact on the History of Civilisation; pp. 77–113. [DOI] [Google Scholar]
  47. Erdkamp P. In: People, land, and politics Demographic developments and the transformation of Roman Italy, 300 BC-AD 14. De Ligt L, Northwood S, editors. BRILL; 2008. Mobility and migration in Italy in the second century BC; pp. 417–449. [Google Scholar]
  48. Ferembach D, Schwidetzky I, Stloukal M. Recommendations for age and sex diagnoses of skeletons. J Hum Evol. 1980;9(7):517–549. doi: 10.1016/0047-2484(80)90061-5. [DOI] [Google Scholar]
  49. Fojas CL. Abandonment of the middle Cumberland region of Tennessee during the Mississippian period: temporal and sex differences in survivorship. Am J Biol Anthropol. 2022;177(3):425–438. doi: 10.1002/ajpa.24446. [DOI] [PubMed] [Google Scholar]
  50. Galanaud P, Galanaud A, Giraudoux P, Labesse H. Mortality and demographic recovery in early post-black death epidemics: role of recent emigrants in medieval Dijon. PLoS One. 2020;15(1):e0226420. doi: 10.1371/journal.pone.0226420. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Galley C. A model of early modern urban demography. Econ Hist Rev. 1995;48(3):448. doi: 10.2307/2598176. [DOI] [Google Scholar]
  52. Gates C. Ancient Cities. Routledge; 2011. Ancient Cities. The Archaeology of Urban Life in the Ancient Near East and Egypt, Greece and Rome. [DOI] [Google Scholar]
  53. Ham AC. Evaluating the effect of sex on mortality risks in medieval Ireland. Am J Biol Anthropol. 2025;186(4):e70040. doi: 10.1002/ajpa.70040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Hanson JW, Ortman SG, Bettencourt LMA, Mazur LC. Urban form, infrastructure and spatial organisation in the Roman Empire. Antiquity. 2019 doi: 10.15184/aqy.2018.192. [DOI] [Google Scholar]
  55. Harper K. The Fate of Rome: Climate, Disease and the End of an Empire. Princeton University Press; 2017. [Google Scholar]
  56. Havlíček F, Morcinek M. Waste and pollution in the ancient Roman Empire. J Landsc Ecol. 2016;9(3):33–49. doi: 10.1515/jlecol-2016-0013. [DOI] [Google Scholar]
  57. Hesketh T, Xing ZW. Abnormal sex ratios in human populations: causes and consequences. Proc Natl Acad Sci. 2006;103(36):13271–13275. doi: 10.1073/pnas.0602203103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Hillson S. The world ‘s largest infant cemetery and its potential for studying growth and development. New Directions in the Skeletal Biology of Greece Hesperia Supplements. 2009;43(2009):137–154. [Google Scholar]
  59. Hooper PL. Forced Population Transfers in Early Ottoman Imperial Strategy: A Comparative Approach. Department of Near Eastern Studies, Princeton University; 2003. [Google Scholar]
  60. Hoppa RD, Vaupel JW. Paleodemography: Age Distributions from Skeletal Samples. Cambridge University Press; 2002. [Google Scholar]
  61. Jones TL, Schwitalla AW, Pilloud MA, Johnson JR, Paine RR, Codding BF. Historic and bioarchaeological evidence supports late onset of post-Columbian epidemics in Native California. Proc Natl Acad Sci U S A. 2021;118(28):e2024802118. doi: 10.1073/pnas.2024802118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Kaiafa-Saropoulou M, Karadedos G. Grammenos VD, editor. H διαχείριση του νερού στην αρχαία Mακεδονία. Στη Mακεδονία από τον 7ο αι πχ Ως την ύστερη αρχαιότητα. 2011:477–500. Zήτρο&#962. [Google Scholar]
  63. Kaminiates J. John Kaminiates—the capture of Thessaloniki. Brill. 2000 doi: 10.1163/9789004344723. [DOI] [Google Scholar]
  64. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53(282):457–481. doi: 10.1080/01621459.1958.10501452. [DOI] [Google Scholar]
  65. Karidas N. Aνασκαφές στην Oδού Γλαύκου. To Aρχαιολογικό Έργο Στη Mακεδονία Kαι Στη Θράκη. 1999:249–258. [Google Scholar]
  66. Karpozilos A. Aggelidi Ch., editor. Περί αποπάτων, βόθρων και υπονόµων. H καθηµερινή ζωή στο Bυζάντιο: Toµές και συνέχειες στην ελληνιστική και ρωµαϊκή παράδοση. 1989:335–352. Kέντρο Bυζαντινών Eρευνών (KBE) / Eθνικό Ίδρυµα Eρευνών (EIE. [Google Scholar]
  67. Kelmelis S, DeWitte SN. Urban and rural survivorship in pre- and post-Black Death Denmark. J Archaeol Sci Rep. 2021;38:103089. doi: 10.1016/J.JASREP.2021.103089. [DOI] [Google Scholar]
  68. Killgrove K, Montgomery J. All roads lead to Rome: exploring human migration to the eternal city through biochemistry of skeletons from two imperial-era cemeteries (1st-3rd c AD) PLoS One. 2016 doi: 10.1371/journal.pone.0147585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Klein Goldewijk K, Beusen A, Janssen P. Long-term dynamic modeling of global population and built-up area in a spatially explicit way: HYDE 3.1. Holocene. 2010 doi: 10.1177/0959683609356587. [DOI] [Google Scholar]
  70. Kleinbaum DG, Klein M. Survival analysis: A Self-Learning text. New York: Springer; 2012. [DOI] [Google Scholar]
  71. Kourkoutidou-Nikolaidou E, Tourta A. Wandering in Byzantine Thessaloniki. Kapon Editions. 1997. [Google Scholar]
  72. Lagia A. Notions of childhood in the classical polis: evidence from the bioarchaeological record. Hesperia Suppl. 2007;41:293–306. [Google Scholar]
  73. Laiou AE, Morrisson C. The Byzantine Economy. Cambridge University Press; 2007. [DOI] [Google Scholar]
  74. Lambrothanassi E, Protopsalti T, Konstantinidou K, Tzevreni S, Vasileiadou S. In: The Metro-nome of Thessaloniki’s History. Adam-Veleni P, Mylopoylos Y, editors. Hellenic Ministry of Culture & Sports, Ephorate of Antiquities of Thessaloniki City, Attiko Metro SA, Hellenic Ministry of Infrastructures & Transport; 2018. The stations. [Google Scholar]
  75. Lancy DF. In: Different faces of attachment. 1st. Otto H, Keller H, editors. Cambridge University Press; 2014. Babies aren’t persons: A survey of delayed person-hood; pp. 66–110. [DOI] [Google Scholar]
  76. Lebegyev J. Ωοκοτο. Child Past. 2009;2(1):15–32. doi: 10.1179/cip.2009.2.1.15. [DOI] [Google Scholar]
  77. Ledger ML, Stock F, Schwaiger H, Knipping M, Brückner H, Ladstätter S, Mitchell PD. Intestinal parasites from public and private latrines and the harbour Canal in Roman period Ephe-sus, Turkey (1st c. BCE to 6th c. CE) J Archaeol Science: Rep. 2018;21:289–297. doi: 10.1016/j.jasrep.2018.07.013. [DOI] [Google Scholar]
  78. Lewis ME. Impact of industrialization: comparative study of child health in four sites from medieval and postmedieval England (A.D. 850–1859) Am J Phys Anthropol. 2002;119(3):211–223. doi: 10.1002/ajpa.10126. [DOI] [PubMed] [Google Scholar]
  79. Liston MA, Rotroff SI. Oxford University Press; 2013. Babies in the well: archeological evidence for newborn disposal in Hellenistic Greece. [DOI] [Google Scholar]
  80. Lovejoy CO, Meindl RS, Pryzbeck TR, Mensforth RP. Chronological metamorphosis of the auricular surface of the ilium: a new method for the determination of adult skeletal age at death. Am J Phys Anthropol. 1985;68(1):15–28. doi: 10.1002/ajpa.1330680103. [DOI] [PubMed] [Google Scholar]
  81. Lowry HW. In: Bryer AAM, Lowry HW, editors. From lesser wars to the mightiest war: The Otto-man conquest and the transformation of Byzantine urban centers in the fifteenth century; Continuity and Change in Late Byzantine and Early Ottoman Society Papers given at a Symposium at Dumbarton Oaks in May 1982; 1986. pp. 323–338. [Google Scholar]
  82. Mackenbach JP. The rise and fall of diseases: reflections on the history of population health in Europe since ca. 1700. Eur J Epidemiol. 2021;36(12):1199–1205. doi: 10.1007/s10654-021-00719-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Marki E. In: Iστορία της επιχειρηµατικότητας στη Θεσσαλονίκη. Adam-Veleni Π, Marki E, editors. Πολιτιστική Eταιρεία Eπιχειρηµατιών B; 2004. H βυζαντινή περίοδος; pp. 153–225. Eλλάδο&#962. [Google Scholar]
  84. Marki E. H Nεκρόπολη της Θεσσαλονίκης στους Yστερορωµαϊκούς και Παλαιοχριστιανικούς χρόνους. Tαµείο Aρχαιολογικών Πόρων και Aπαλλοτριώσεων, Διεύθυνση Δηµοσιευµάτων. 2006 [Google Scholar]
  85. Matschke KP. In: The economic history of Byzantium From the seventh through the fifteenth century Dumbarton Oaks Research Library and Collection. Laiou AE, editor. 2002. The late Byzantium urban economy, thirteenth-fifteenth centuries; pp. 463–495. [Google Scholar]
  86. McFadden C. The past, present and future of skeletal analysis in palaeodemography. Philos Trans R Soc Lond B Biol Sci. 2021;376(1816):20190709. doi: 10.1098/rstb.2019.0709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  87. McFadden C, Oxenham MF. The D0-14/D ratio: a new paleodemographic index and equation for estimating total fertility rates. Am J Phys Anthropol. 2018;165(3):471–479. doi: 10.1002/ajpa.23365. [DOI] [PubMed] [Google Scholar]
  88. Meindl RS, Lovejoy CO. Ectocranial suture closure: a revised method for the determination of skeletal age at death based on the lateral-anterior sutures. Am J Phys Anthropol. 1985 doi: 10.1002/ajpa.1330680106. [DOI] [PubMed] [Google Scholar]
  89. Mitchell PD. Human parasites in the Roman world: health consequences of conquering an empire. Parasitology. 2017;144(1):48–58. doi: 10.1017/S0031182015001651. [DOI] [PubMed] [Google Scholar]
  90. Moore DF. Applied survival analysis using R. Springer Int Publishing; 2016. [DOI] [Google Scholar]
  91. Mueller-Scheessel N, Hinz M, Schmid C. mortAAR: Analysis of Archaeological Mortality Data. 2025. https://cran.r-project.org/web/packages/mortAAR .
  92. Müller-Scheeßel N, Rinne C, Fuchs K. Adult mortality in the metropolis of London 1100–1850: a Bayesian view based on osteological data. Am J Biol Anthropol. 2024;185(3):e25025. doi: 10.1002/ajpa.25025. [DOI] [PubMed] [Google Scholar]
  93. Nagaoka T, Ishida H, Tsurumoto T, Wakebe T, Saiki K, Hirata K. A health crisis during the Japanese medieval period: a new paleodemographic perspective. Int J Paleopathol. 2019;26:145–156. doi: 10.1016/j.ijpp.2019.03.005. [DOI] [PubMed] [Google Scholar]
  94. Newton G, Smith R. Convergence or divergence? Mortality in London, its suburbs and its hinterland between 1550 and 1700. Ann De Démographie Historique. 2013;126(2):17–49. doi: 10.3917/adh.126.0017. [DOI] [Google Scholar]
  95. Nigdelis P. In: From Roman to Early Christian Thessalonikē Studies in Religion and Archaeology. Nasrallah L, Bakirtzis C, editors. Harvard theological studies; 2010. Voluntary Associations in Roman Thessalonikē: In Search of Identity and Support in a Cosmopolitan Society; pp. 13–48. [Google Scholar]
  96. Novak M. Health and disease in a Roman walled city: an example of colonia Iulia Iader. J Anthropol Sci. 2010;88:189–206. [PubMed] [Google Scholar]
  97. Olga Hernández Espinoza P, Márquez Morfín L. Maya paleodemographics: what do we know? Am J Hum Biol. 2015;27(6):747–757. doi: 10.1002/ajhb.22751. [DOI] [PubMed] [Google Scholar]
  98. Paine RR. Model life table fitting by maximum likelihood estimation: a procedure to reconstruct paleodemographic characteristics from skeletal age distributions. Am J Phys Anthropol. 1989;79(1):51–61. doi: 10.1002/ajpa.1330790106. [DOI] [PubMed] [Google Scholar]
  99. Phenice TW. A newly developed visual method of sexing the Os pubis. Am J Phys Anthropol. 1969;30(2):297–301. doi: 10.1002/ajpa.1330300214. [DOI] [PubMed] [Google Scholar]
  100. Prowse TL. Migration and Mobility in the Early Roman Empire. BRILL; 2016. Isotopes and Mobility in the Ancient Roman World; pp. 205–233. [DOI] [Google Scholar]
  101. Reba M, Reitsma F, Seto KC. Spatializing 6,000 years of global urbanization from 3700 BC to AD 2000. Sci Data. 2016 doi: 10.1038/sdata.2016.34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  102. Redfern RC, Dewitte SN. A new approach to the study of Romanization in Britain: a regional perspective of cultural change in late Iron Age and Roman Dorset using the Siler and Gompertz-Makeham models of mortality. Am J Phys Anthropol. 2011 doi: 10.1002/ajpa.21400. [DOI] [PMC free article] [PubMed] [Google Scholar]
  103. Redfern RC, Dewitte SN, Pearce J, Hamlin C, Dinwiddy KE. Urban-rural differences in Roman Dorset, England: a bioarchaeological perspective on Roman settlements. Am J Phys Anthropol. 2015;157(1):107–120. doi: 10.1002/ajpa.22693. [DOI] [PubMed] [Google Scholar]
  104. Reyes R, Ahn R, Thurber K, Burke TF. In: Challenges in Infectious Diseases. Fong IW, editor. Springer New York; 2013. Urbanization and Infectious Diseases: General Principles, Historical Perspectives, and Contemporary Challenges; pp. 123–146. [DOI] [Google Scholar]
  105. Roberts CA, Manchester K. The archaeology of disease. History. 2010 [Google Scholar]
  106. Salomies O. In: Rizakis A, editor. Contacts between Italy, Macedonia and Asia Minor during the Principate; Roma Onomastics in the Greek East: Social and Political Aspects: Proceedings of the International Colloquium on Roman Onomastics Athens 7–9 September 1993; De Boccard; 1996. pp. 118–123. [Google Scholar]
  107. Sammut S. Reconstructing health on the Danube Limes: evidence from eastern Austria. J Archaeol Sci Rep. 2025;61:104888. doi: 10.1016/j.jasrep.2024.104888. [DOI] [Google Scholar]
  108. Sattenspiel L, Harpending H. Stable populations and skeletal age. Am Antiq. 1983;48(3):489–498. doi: 10.2307/280557. [DOI] [Google Scholar]
  109. Saunders SR, Barrans L. In: Human growth in the past: studies from bones and teeth. Hoppa RD, Fitzgerald CM, editors. Cambridge University Press; 1999. What can be done about the infant category in skeletal samples; pp. 183–209. [Google Scholar]
  110. Sharlin A. Natural decrease in early modern cities: a reconsideration. Past Present. 1978;79(1):126–138. doi: 10.1093/past/79.1.126. [DOI] [Google Scholar]
  111. Smith ME, Lobo J, Peeples MA, York AM, Stanley BW, Crawford KA, Gauthier N, Huster AC. The persistence of ancient settlements and urban sustainability. Proc Natl Acad Sci U S A. 2021;118(20):e2018155118. doi: 10.1073/pnas.2018155118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  112. Stathakopoulos D. The Justinianic plague revisited. Byz Mod Greek Stud. 2000;24(1):255–276. doi: 10.1179/byz.2000.24.1.255. [DOI] [Google Scholar]
  113. Stefanidou-Tiveriou T. In: From Roman to early Christian Thessalonikē studies in religion and archaeology. Nasrallah L, Bakirtzis C, Friesen SJ, editors. Harvard University Press; 2010. Social status and family origin in the sarcophagi of Thessalonikē; pp. 151–188. [Google Scholar]
  114. Stojanowski CM, Seidemann RM, Doran GH. Differential skeletal preservation at Windover Pond: causes and consequences. Am J Phys Anthropol. 2002;119(1):15–26. doi: 10.1002/ajpa.10101. [DOI] [PubMed] [Google Scholar]
  115. Stoline MR, Ury HK. Tables of the studentized maximum modulus distribution and an application to multiple comparisons among means. Technometrics. 1979;21(1):87–93. [Google Scholar]
  116. Storey R. An estimate of mortality in a pre-Columbian urban population. Am Anthropol. 1985;87(3):519–535. [Google Scholar]
  117. Tacoma LE. Moving romans: migration to Rome in the principate. Oxford University Press; 2016. [DOI] [Google Scholar]
  118. Taylor BR, Oxenham MF. A method for detecting bias in human archaeological cemetery samples. Int J Osteoarchaeol. 2025;35(1):e3379. doi: 10.1002/oa.3379. [DOI] [Google Scholar]
  119. Therneau TM. A Package for Survival Analysis in R. 2024 [Computer software] https://CRAN.R-project.org/package=survival .
  120. Therneau TM, Grambsch PM. Modeling survival data: extending the Cox model. Springer; 2000. [Google Scholar]
  121. Torres C, Canudas-Romo V, Oeppen J. The contribution of urbanization to changes in life expectancy in Scotland, 1861–1910. Popul Stud. 2019;73(3):387–404. doi: 10.1080/00324728.2018.1549746. [DOI] [PubMed] [Google Scholar]
  122. Treadgold W. A history of the Byzantine state and society. Stanford University Press; 1997. [Google Scholar]
  123. Tsiamis C, Poulakou-Rebelakou E, Tsakris A, Petridou E. Epidemic waves of the black death in the Byzantine empire (1347–1453 AD) Le Infezioni in Medicina. 2011;19(3):194–201. [PubMed] [Google Scholar]
  124. United Nations. World Urbanization Prospects 2018. Department of Economic and Social Affairs; 2018. World Population Prospects 2018. [Google Scholar]
  125. Vacalopoulos A. History of Macedonia 1354–1833. Institute for Balkan Studies; 1973. [Google Scholar]
  126. Vanderspoel J. In: A companion to ancient Macedonia. Roisman J, Worthington I, editors. Wiley-Blackwell; 2010. Provincia Macedonia; pp. 251–276. [Google Scholar]
  127. Vitti M. From the foundation to Galerius. The Archaeological Society at Athens; 1996. The urban planning of Thessaloniki. [Google Scholar]
  128. Vokotopoulou I. Aνασκαφή Mένδης 1989. To Aρχαιολογικό Έργο Στη Mακεδονία Kαι Στη Θράκη. 1989;3:409–423. [Google Scholar]
  129. Walter BS, DeWitte SN. Urban and rural mortality and survival in medieval England. Ann Hum Biol. 2017;44(4):338–348. doi: 10.1080/03014460.2016.1275792. [DOI] [PubMed] [Google Scholar]
  130. Wataghin GC. The Idea and Ideal of the Town between Late Antiquity and the Early Middle Ages. BRILL; 1999. The Ideology of Urban Burials; pp. 147–180. [DOI] [Google Scholar]
  131. Weiss KM. On the systematic bias in skeletal sexing. Am J Phys Anthropol. 1972;37(2):239–249. doi: 10.1002/ajpa.1330370208. [DOI] [PubMed] [Google Scholar]
  132. Weiss KM, Wobst MH. Demographic models for anthropology. Mem Soc Am Archaeol. 1973;27:i-186. [Google Scholar]
  133. White SD, Mahoney P, Deter CA. Socioeconomic status and survival in medieval Canterbury. J Archaeol Sci Rep. 2022;46:103686. doi: 10.1016/j.jasrep.2022.103686. [DOI] [Google Scholar]
  134. Woods R. Urban-rural mortality differentials: an unresolved debate. Popul Dev Rev. 2003;29(1):29–46. [Google Scholar]
  135. Yaussy SL, DeWitte SN, Hughes-Morey G. Survivorship and the second epidemiological transition in industrial-era London. Am J Biol Anthropol. 2023;181(4):646–652. doi: 10.1002/ajpa.24797. [DOI] [PubMed] [Google Scholar]
  136. Zarulli V, Barthold Jones JA, Oksuzyan A, Lindahl-Jacobsen R, Christensen K, Vaupel JW. Women live longer than men even during severe famines and epidemics. Proc Natl Acad Sci. 2018;115(4) doi: 10.1073/pnas.1701535115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  137. Zoeller GE, Drew BL, Schmidt CW, Peterson R, Wilson JJ. A paleodemographic assessment of mortality and fertility rates during the second demographic transition in rural central Indiana. Am J Hum Biology. 2022;34(1) doi: 10.1002/ajhb.23571. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information
EMS217082-supplement-Supplementary_Information.docx (57.5KB, docx)

Data Availability Statement

This article and the SI include all data generated and analyzed during this study.

RESOURCES