The human remains of early rice farmers used as materials in this study included eight burial human skeletons from the Hemudu site and 14 from the Tianluoshan site, with a total of 22 burial skeletons (Table 1). In addition, 12 scattered human remains excavated from the Tianluoshan site were used as materials (mentioned later). The Hemudu and Tianluoshan sites are located in Zhejiang Province in southern China, and both are from the Hemudu culture–7,000-5,500 years ago (Figure 1).

The Hemudu site was discovered in 1973 and excavations were conducted between 1973 and 1978 (Zhejiang, 2003). Eight human skeletons (one adult female, two adult males, and five juveniles of unknown sex) were found during the excavations and stored at the Hemudu Site Museum. The skulls of one adult male and one adult female were cleaned, and the results of the morphological research have been reported (Zhejiang, 2003). However, the post-cranial trunk and limb parts of these two skeletons were not exhumed, and the entire grave pits were preserved in the museum. The other six whole skeletons were not exhumed; all grave pits were preserved in the museum, and their anthropological reports have not been published. In this study, we examined eight human skeletons for CO, LEH, dental caries, and AMTL. However, some parts of the skull that were in an earth-encased state could not be adequately observed. In addition, the postcranial parts were not investigated for periosteal reactions because it was impossible to remove them from the soil of the grave pit.

The Tianluoshan site, located 7 km from the Hemudu site, was discovered in 2001 and has been continuously excavated since 2004. Its archaeological results have been reported by Nakamura (2010b) and Matsui and Kikuchi (2016); however, no anthropological reports of excavated human remains have been published. In this study, 14 buried human skeletons (four adult females, five adult males, and five juveniles of unknown sex) stored at the Tianluoshan site museum, and 12 scattered human remains (one adult mandible, two adult right femurs, one adult left femur, one adult left tibia, and seven permanent teeth) that were considered to have originated from individuals other than those in these burials, were examined for the presence of CO, LEH, periosteal reaction, dental caries, and AMTL.

There were no hunter-gatherer sites in the lower reaches of the Yangtze River, from which many human remains were excavated. Therefore, the human skeletons of non-agricultural hunter-gatherers in the Early Holocene, approximately 9,000 to 7,000 years ago, from the Huiyaotian and Liyupo sites in Guangxi Province, Southern China (Matsumura et al., 2017), were used as comparative materials for early farmers in this study (Figure 1). The Huiyaotian and Liyupo sites, as well as the Hemudu and Tianluoshan sites, belong to the warm humid climate (Cfa) category of southern East Asia in the Köppen climate classification. They share similarities with the Hemudu cultural sites in being located in a plain area along the lower reaches of a large river. Paleopathological investigations of 91 hunter-gatherer human skeletons from the Huiyaotian and Liyupo sites were conducted by one of the authors of this study (JS) using the same methods as those used in this study (Sawada, 2017a; Sawada, 2017b; Sawada et al., 2017).

The age of juvenile human remains was estimated based on the tooth formation status (Smith, 1991), eruption status (Ubelaker, 1999), and the fusion status of the epiphyseal ends of the limb bones (Maresh, 1970; Fazekas and Kosa, 1978; Cunningham et al., 2016). The age of adult remains was estimated based on the closure status of the cranial sutures (White et al., 2012) and the metamorphosis of the pubic symphysis surface (Brooks and Suchey, 1990).

The sex of the human remains was estimated from the shape of the adult coxal bone and skull, based on the sex determination method in White et al. (2012).

Fisher’s exact test was used to examine any statistically significant difference in the frequency of occurrence of each paleopathological characteristic between early farmers and hunter-gatherers. The significance level was set at p < 0.05. Statistical analyses were conducted using the SPSS software package for Macintosh, version 25.0.0 (IBM).

The occurrence ratio of CO in juveniles of early rice farmers was 25.0%, and that in adults was 0.0% (Table 2). By contrast, the incidence of CO among hunter-gatherers was 12.5% in adults (no juveniles with preserved supraorbital walls). Fisher’s exact test was used to examine any statistical difference in the frequency of CO occurrence between farmers and hunter-gatherers. No significant difference was found at the 0.05 level.

The occurrence rates of LEH in early rice farmers in UI1 and LC were 80.0% and 85.7%, respectively. The occurrence rates of LEH in the hunter-gatherer population were 73.7% and 50% in the UI1 and LC, respectively (Table 3). Fisher’s exact test showed no statistically significant differences between farmers and hunter-gatherers at the 0.05 level.

As mentioned above, examining the limb bones from the Hemudu site was impossible; therefore, only the bones from the Tianluoshan site were used to analyze the periosteal reaction. Periosteal reactions observed in these materials were mild. The occurrence ratio of periosteal reactions in the adult lower limb long bones of the farmers (19.0%) tended to be higher than that of the hunter-gatherers (2.7%), and Fisher’s exact test showed a statistically significant difference (Table 4).

The percentage of dental caries in the total number of erupted permanent teeth was 12.5% in early rice farmers compared to 5.5% in hunter-gatherers, a statistically significant difference at the 0.05 level in Fisher’s exact test (Table 5). The dental caries ratio of the Huiyaotian and Liyupo hunter-gatherers used for comparison in this study tended to be similar to the low caries rates of modern hunter-gatherer groups (Turner, 1979).

The occurrence rate of AMTL in the early rice farmers, calculated based on the number of permanent tooth sockets, was 7.8%. This tended to be higher than that in hunter-gatherers (4.8%), although no statistically significant difference was detected (Table 6).

No statistically significant difference in the frequencies of CO and LEH were found between early rice farmers of the Hemudu culture and hunter-gatherer groups in southern China. Although our findings are preliminary owing to the small sample size of early rice farmers, there was no evidence that the health status of early rice farmers was appreciably different from that of hunter-gatherers. While it is thought that health declined after the introduction of agriculture in North America and the Mediterranean (Goodman et al., 1980; Angel, 1984; Smith et al., 1984; Larsen, 1995), such changes may not have occurred in the Neolithic societies of East Asia that began rice farming. In this regard, examining the characteristics of people’s livelihoods during Hemudu culture may be necessary. Zooarchaeological studies suggest that the people of the Hemudu culture consumed a variety of mammals and fish acquired through hunting and fishing, as well as domestic animals, such as pigs and geese (Matsui and Kikuchi 2003; Eda et al., 2022). A diversified food acquisition strategy that did not rely heavily on rice farming may have contributed to a stable food supply and prevented a significant decline in the health status of the people during the Hemudu culture period.

However, for periosteal reactions in adult bone, the occurrence ratio in the early rice farmers was significantly higher than that in the hunter-gatherer group (Figure 2). The periosteal reaction is an early response of the bone to the effects of disease or trauma, and is considered an indicator of non-specific stress (Larsen, 1999; Waldron, 2021). Although it is difficult to extrapolate a specific causative lesion from the periosteal reaction alone (Weston, 2008), unlike LEH, which is derived from stress in infancy, the periosteal reaction in adult bones is probably derived from stress at a later age. The high incidence of periosteal reactions in early rice farmers may be related to infections caused by long hours of farming in rice fields with high mechanical stress, or noxious insects such as schistosomes (Ross et al., 2013) and mosquitoes that are easily contacted in the aquatic environment. Future studies on human bones for paleopathological features associated with infection and whether DNA from infection remains in human bones will allow for a more specific discussion of the etiology of periosteal reactions.

The occurrence ratios of dental caries and AMTL in people of the Hemudu culture tended to be higher than those in the hunter-gatherer group (Figure 2). These results suggest that early rice-farming societies in the lower reaches of the Yangtze River had factors that contributed to oral hygiene deterioration. In general, the occurrence of dental caries is related to factors such as caries pathogens (e.g., mutans), dietary characteristics (sugar and sticky foods), dental habits, and lifestyle (Hillson, 1996; Hillson, 2008; Waldron, 2021). Temple and Larsen (2007) reported that a rice-based diet is responsible for dental caries in rice-farming populations. However, Willis and Oxenham (2013), who presented data on the decline in oral health during the introduction of agriculture in Southeast Asia, are critical of the view that rice is the main cause of dental caries, based on a review by WHO (2003), which stated that “rice has low potential cariogenicity.” Willis and Oxenham (2013) highlighted sexual differences in the quality of oral hygiene and physiological factors that lead to poor oral hygiene in females (especially during pregnancy).

We could not examine sexual differences because the sample size was too small to allow sex determination owing to the large number of juvenile and fragmentary human bone materials that are difficult to determine. We hope that future excavations will increase the amount of human material from early farming groups in the lower reaches of the Yangtze River and advance the study of sexual differences in the occurrence ratios of paleopathological features.