Agriculture and Dental Caries? The Case of Rice in Prehistoric Southeast Asia Author(s): N. Tayles, K. Domett, K.

Nelsen Reviewed work(s): Source: World Archaeology, Vol. 32, No. 1, Archaeology in Southeast Asia (Jun., 2000), pp. 6883 Published by: Taylor & Francis, Ltd. Stable URL: http://www.jstor.org/stable/125047 . Accessed: 20/12/2011 17:15
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to World Archaeology.

http://www.jstor.org

Agriculture
Southeast

and Asia

dental

caries?

The case of rice in prehistoric
N. Tayles, K. Domett and K. Nelsen

Abstract
The agricultural transition has long been recognized to have been a very important period in human prehistory. Its timing and consequences, including the effects on human health, have been intensively researched. In recent decades, this has included the idea that there is a universal positive correlation between the adoption of agriculture based on a carbohydrate staple crop and dental caries prevalence. This is mainly based on evidence from America, where maize was the staple crop. On the basis of evidence from prehistoric skeletal samples from a series of prehistoric sites in Southeast Asia, this correlation does not appear to apply in areas of the world where the staple crop is rice. Although we have looked only at dental caries, we suggest that caution be applied in the drawing of inferences about subsistence changes from dental health. Patterns reflecting the adoption of one starchy staple are not necessarily applicable to all such crops.

Keywords
Rice; agriculture; dental caries; prehistory; Southeast Asia.

Introduction

The adoption of agricultural economies, such as those based on the cultivation of dornesticated starchy crops, represents a major change in subsistence mode in prehistory. In the past, estimates of the timing of this important transition relied on indirect evidence in the form of cultivating tools or cooking pots. More recently, the development of techniques for the recovery of botanical remains from archaeological excavations has provided direct evidence of cultivated or domesticated plants. However, in the absence of this evidence, particularly where results from older excavations are being reworked, a further alternative is the remains of the people themselves. The prevalence and patterns of dental pathology, especially dental caries, are particularly favoured because of the direct effect of diet and the high level of preservation of dental enamel. Increasingly, the idea has been adopted that
LCao

World Archaeology

Vol. 32(1): 68-83

Archaeology

in Southeast Asia

Arcaeoogy l

? Taylor & Francis Ltd 2000 0043-8243

Agriculture and dental caries? 69 the development of agricultureresulted in a higher prevalence of dental caries (Molnar and Molnar 1985: 60; Lubell et al. 1994: 211; Larsen 1997: 70). This is based on the assumptions that, first, the development of agriculture resulted in a diet with a high proportion of a carbohydrate staple (particularly grains such as maize, wheat and rice, but also roots or tubers such as taro and yams) and, second, that all carbohydrates are cariogenic. Publications (e.g. Larsen 1982, 1983) which have suggested a positive correlation between the adoption of agriculture and rates of dental caries in Native American populations are based on an increase in caries rates with the adoption of maize agriculture. In association with the overall increase in dental caries with an agricultural lifestyle, there are suggestions that there was also an increased sex difference in the caries rates, with higher rates in females than in males in most regions (Larsen 1983, 1995; Larsen et al. 1991). Beyond North America, an influential paper by Turner (1979) considered the topic in the context of research into the subsistence mode of the prehistoric Jomon people of Japan. His paper went far beyond Japan and collated a vast amount of data on caries rates in populations, both prehistoric and modern, with different subsistence modes. On the basis of the mean caries rates in each of the categories of hunting and gathering, mixed and agricultural economies, he suggested the increase in caries rates was applicable world-wide. In subsequent years the relationship between dental pathology, including caries, and the adoption of agriculturehas been further investigated and shown to be positively correlated elsewhere in the world (e.g. Lukacs 1992). The question was included in a symposium on the effects on human health of the development of agriculture (Cohen and Armelagos 1984). This included papers on populations from a wide range of countries and continents, with different subsistence modes, but did not include any populations with rice as their carbohydrate staple, nor have any more recent publications specifically considered the issue. We therefore set out here to test the hypothesis that the transition to an agricultural economy based on rice resulted in an increase in dental caries. For this we use skeletal remains from three populations from prehistoric Southeast Asia, spanning the periods from the pre-metal era through to the development of rice-based agricultural groups using first bronze, and later iron, technology. Domesticated rice has been part of the diet in at least some areas of Southeast Asia from at least 2000 BC (Thompson 1996). It would have been an important component of the diet by the Bronze Age -1000 BC and agriculture would have intensified as a subsistence mode by the Iron Age -500 BC. In addition, this paper presents the first stage of publication of our broader research topic of determining the variation over time in health status of people living in prehistoric Southeast Asia. We recognize that dental health is much more complex than can be demonstrated by consideration of dental caries in isolation from the consequences of tooth wear and the deterioration of the periodontal tissues. These data, together with other evidence of responses to dietary change, will be presented in future publications.

Materials The three sites we are using are all from the modern state of Thailand. The site of Khok Phanom Di is in central Thailand, near the Bang Pakong River, and Ban Lum Khao and

70 N. Tayles,K. Domett and K. Nelsen Noen U-Loke are in the Mun River Valley, on the Khorat Plateau of North East Thailand. The earliest site, Khok Phanom Di, is most appropriately described as a rich sedentary hunter-gatherer site. It is near the coast of the Gulf of Thailand (Fig. 1). An excavation in 1985 recovered 154 virtually complete and well-preserved skeletons. The sample used in this paper includes sixty-seven (of sixty-eight) adults and ten (of twelve) children with permanent teeth (Table 1). Radiocarbon dates suggest the cemetery deposits covered about 500 years, from cal. 2000-1500 BC (Higham and Bannanurag 1990). Detailed archaeobotanical and palynological studies from the site and environs show that the diet included domesticated rice, although, as the immediate vicinity of the site (at the time near the coast) was probably too saline for its cultivation, the rice was clearly being transported in from nearby. Evidence from modern tropical coastal economies suggests that the diet would have been broad spectrum, including many wild plant foods (Thompson 1996). A variety of plants which grow in the vicinity of the site at present, including the important coconut palm, Cocos nucifera, and other fruit trees (Maloney 1991; Thompson 1996) would have contributed to the diet. Food debris indicates that the people also ate

. . . .

. .
R.

.I. e

.

:. ^j. i

:; 0::: .',,/' .:- ,

: 1 'i :: :,'' ^,:::::/ ^ :,

0

Jjf |llll8 Si .

igi :0:South Gulf of Thailand 2

,

China Sea

showing
-

the sites referred to in

.,,,,,||f,
0

:?|
300 0 km
km

, : I;: :;::;:;'''the
* e?

\,Mj^ I.i

text. 1 Figure 1 Map of Southeast Asia, Figure of showirng the sites referred to in
the text.

--0

100

2,00

200

l?-/^

South China Sea

Agriculture and dental caries? 71 very large numbers of shellfish, together with fish and crustaceans, and a relatively limited amount of meat from terrestrial animals. One of the Northeastern Thai sites, Ban Lum Khao, is Bronze Age and the second, Noen U-Loke, is Iron Age. They are within 20km of one another, in similar environments on a river channel (Fig. 1). Ban Lum Khao is a modern village built on a prehistoric mound. An excavation in 1996 recovered 110 individuals from a Bronze Age cemetery between two Bronze age occupation levels. Not all skeletons were complete, but there were forty-three adults and eight children with permanent teeth present used in this paper (Table 1). Dating has indicated the initial occupation at around 1400 BC, with the cemetery phase being dated between approximately 1000 and 500 BC (Higham and Thosarat 1998). No results are yet available from research on botanical material collected and recovered from flotation, but preliminary analyses of the temper from a small sample of potsherds show that rice was abundant in the upper levels of the site. The evidence from potsherds from the lower levels, which include most of the burials, is as yet unclear but does not exclude the possibility of the use of rice chaff as temper (J. Voelker pers. comm.). This suggests that rice was part of the diet of the people buried in the cemetery. Organic remains included many fish and shellfish and a variety of mammalian bones, both wild and domesticated. This included water buffalo, pig and deer (Higham and Thosarat 1998). Noen U-Loke is a large mound, now surrounded by paddy fields. It is one of numerous 'moated' sites in the valley of the Mun River, dating from the Iron Age. These appear to have been substantial villages, sited on relatively high areas on the banks of rivers. Excavations at Noen U-Loke over two seasons between 1996 and 1998 recovered a sample of 127 burials, including forty-two adults and four children with permanent teeth used in this paper (Table 1). The bone in many burials was very poorly preserved but the teeth are largely intact. Radiocarbon dates show that the cemetery was in use during the period from c. 300 BC to c. AD 300 (Higham and Thosarat 1998). Little information is available on the prehistoric environment and diet as yet, although an unusual aspect of the burial ritual for a period during the cemetery use was the interment of bodies in graves filled with rice. These very large quantities of rice, together with other deposits found in pits, suggest a very productive agricultural economy. There is no reason to believe that this was not supplemented by collecting of wild plants, together with use of freshwater fish and molluscs, and domesticated animals (Higham pers. comm.). Secondary evidence of agriculture is in the form of iron sickles, numerous hoes and a spade which were buried with some of the dead. Work by Boyd (pers. comm.) has shown that the 'moats' are actually river channels which have been modified over time, including up until the present, to Table1 The skeletalsamples
Site Dates Period Location n individuals n with permanent teeth**

Neolithic KhokPhanomDi 2000-1500BC coastal Ban LumKhao 1000-500BC* BronzeAge inland 300 BC-AD 300* IronAge inland Noen U-Loke to *Dates given correspond the cemeteryphase ** Excludingadultswith no age and/orsex estimates

154 110 128

77 51 46

72 N. Tayles,K. Domett and K. Nelsen create paddy fields on the rich alluvial soils deposited as sediment by the rivers. This is therefore further secondary evidence of an intensification of agriculture during the Iron Age. The density of 'moated' sites dating to the Iron Age in the Mun River Valley suggests an increased population in the area, which is also consistent with the development of increased rice production.

Methods Sex of adults aged over 15 years was estimated through standard morphological observations of the pelvis and cranium and some metric analyses. Particularly at Noen U-Loke, where the preservation was poor, the sex of a number of individuals could not be estimated. As identifying sex differences in caries prevalence was one of our objectives, these individuals were excluded from the analysis. A multifactorial approach was taken in estimating the ages of individuals from each of the three samples, although the variable condition of the skeletal material among the samples influenced the degree of accuracy obtainable. Much of the material from Noen U-Loke was fragmentary, with the ages of some adults unable to be estimated; these were also excluded from this analysis. The age structure of the samples is an important factor in the investigation of agerelated dental pathologies. The sex ratio of each sample will also be considered as males and females may be differentially affected by dental disease through differences in diet. To take account of both age and sex, males and females in each sample are analysed separately. The age structure of each group of males and females is compared statistically among the three samples. Subadults were considered separately from the adults, with only those teeth that were permanent and fully erupted included. Only the molars, premolars and canines were analysed. The lateral and central incisors were excluded from all samples as within the Khok Phanom Di sample there was widespread evidence of deliberate ablation of these teeth (Tayles 1996). In the Noen U-Loke sample there were also many individuals with missing incisors, but without the symmetry seen at Khok Phanom Di. The reason for their absence is the subject of current research, with possible explanations being congenital absence and/or ablation. In these circumstances, we have excluded them from our calculations for all three samples. As the incisors are relatively infrequently affected by caries, this should not have a significant effect on the results (Nikiforuk 1985). Caries were identified macroscopically as a necrotic cavity in the tooth crown or root. Every permanent tooth present and visible was assessed. Lesions that had destroyed most of the crown, including the dentine, were noted as 'massive' caries. The observed caries rate was calculated as a simple proportion of the number of carious teeth (or individuals) over the total number of teeth observed (or individuals). A modification to the assessment of the observed caries rate, as proposed by Lukacs (1995), has been adopted here as it takes into account the probable number of teeth lost ante-mortem through caries. Ante-mortem tooth loss was recorded for every tooth position possible, with evidence of remodelling of the alveolar bone surrounding the socket being taken as a positive sign of a tooth lost during life. Teeth may be lost ante-mortem as a result not only of caries, but also of extreme wear which has exposed the pulp cavity and

Agriculture and dental caries? 73 led to infection and ultimately loss or deliberate removal of the tooth. In older individuals, loss can also occur as a result of chronic periodontal disease, loss of support for the tooth, and therefore its exfoliation. To estimate the proportion of teeth lost through caries, the number of caries that exposed the pulp is required. This was taken as equivalent to the massive caries in this study. The total number of teeth with pulp exposure is also required and this is obtained from the number of teeth with massive caries plus the number of teeth with advanced attrition. Attrition was graded for every tooth present using the scale of 1 to 8 proposed by Molnar (1971). Advanced attrition in this study is analogous to a grade of 6 to 8 on the Molnar scale, indicating pulp exposure. The probable proportion of teeth lost by caries can then be estimated (Lukacs 1995). Chi2statistical tests were used when comparing all three samples and a p-value of less than 0.05 was deemed significant. When the Chi2statistical test indicated that a statistical difference existed among the three samples, Fisher's Exact Tests (FET) were performed on all combinations of two samples to determine between which samples the differences exist, with a p-value of less than 0.01 deemed significant.

Results Statistical tests of the age distributions (Table 2) between the male and female samples both within and among the sites show no statistically significant differences. This indicates that, even though Ban Lum Khao and Noen U-Loke appear to have imbalances between the sexes in the subsamples for each age group, the differences should not have affected the results. Table 3 provides the data on massive caries and extreme wear which form the basis of calculation of the corrected caries rates. Corrected caries rates vary among the samples from as high as 21 per cent in the Khok Phanom Di females to as low as 1.9 per cent in the males from Ban Lum Khao (Table 4). There are significant differences between the rates in males and females in two of the samples (Khok Phanom Di and Ban Lum Khao), therefore in all analyses male and female data are considered separately. Female corrected caries rates vary significantly among the three samples (Table 5, Fig.

Table2 Numberof individuals from each site with permanentand fully eruptedteeth in each age groupand sex
Age Khok Phanom Di Ban Lum Khao Noen U-Loke

Subadults (<15 y.)
15-29 y. > 30y. Total adults Male 15 16 31

10
Female 15 21 36 Male 6 13 19

8
Female 15 9 24 Male 15 8 23

4
Female 6 13 19

p-value* Total

0.6285 67

0.0666 43

0.0616 42

males v. females;Fisher'sexact test. *Intrasite, males chi2= 4.729,p value 0.0949;femaleschi2= 4.494,p value 0.1057 Intersite:

74

N. Tayles, K. Domett and K. Nelsen

Table 3 Rates of massive caries, advanced wear and ante-mortem tooth loss used in the calculation of the 'corrected' caries rate (teeth are the unit of measurement) Massive caries n/total Khok Phanom Di Ban Lum Khao Noen U-Loke Khok Phanom Di Ban Lum Khao Noen U-Loke Khok Phanom Di Ban Lum Khao Noen U-Loke 11/540 3/400 6/331 28/557 15/445 1/281 39/1097 18/845 7/612 % 2.0 0.8 1.8 5.0 3.4 0.4 3.6 2.1 1.1 Advanced wear n/total 70/560 70/331 35/338 % Males 12.5 21.1 10.4 Ante-mortem tooth loss n/total 29/715 14/413 10/411 119/829 27/459 24/347 148/1544 44/872 34/758 % 4.1 4.1 2.4 14.4 5.9 6.9 9.6 5.0 4.5

Females 4.6 30/650 5.7 20/349 22.9 64/280 Total 100/1210 90/680 99/618 8.3 13.2 16.0

Table 4 Observed caries rates at all sites using individuals and teeth as the units of measurement and corrected caries rates using teeth as the units of measurement Observed caries (individuals) n/total Khok Phanom Di Ban Lum Khao Noen U-Loke Khok Phanom Di Ban Lum Khao Noen U-Loke Khok Phanom Di Ban Lum Khao Noen U-Loke 15/31 5/19 7/23 24/36 14/24 6/19 39/67 19/43 13/42 % 48.4 26.3 30.4 66.7 58.3 31.6 58.2 44.2 31.0 Observed caries (teeth) n/total 40/540 7/400 14/331 % Males 7.4 1.8 4.2 44/569 8/417 16/341 142/676 42/472 9/305 167/1245 44/889 25/646 7.7 1.9 4.7 21.0 8.9 3.0 13.4 4.9 3.9 Corrected caries rate (teeth) n/total %

Females 85/557 15.3 30/445 6.7 3.2 9/281 125/1097 37/845 23/612 Total 11.4 4.4 3.8

2). The results show a marked decrease with advancing time. Khok Phanom Di females have a caries rate of 21 per cent, over twice as high as the Ban Lum Khao females (8.9 per cent) and seven times higher than the Noen U-Loke females (3 per cent). The caries rates for the male samples follow a similar trend to the female samples but the decrease in caries with time is not as marked. Males from the Neolithic sample of Khok Phanom Di show the highest caries rate (7.7 per cent) followed by the Iron Age males of Noen ULoke with 4.7 per cent. Ban Lum Khao males from the Bronze Age have only 1.9 per cent

Agriculture and dental caries? 75 Table5 Chi2andFisher'sexacttests for significance amongthe correctedcariesrateslistedin Table
4 Chi2 statistic p-value Fisher's exact tests:p-values KPD vs BLK KPD vs NUL BLK vs NUL

Males Females

16.910 70.149

0.0002* <0.0001*

<0.0001* <0.0001'

0.0969 <0.0001*

0.0366 0.0009*

* statistically KPD:Khok PhanomDi. BLK:Ban Lum Khao.NUL: Noen U-Loke significant. caries rate. Among the males, statistical differences exist only between the Khok Phanom Di and Ban Lum Khao samples. Comparisons of the caries rates in male and female skeletons within each sample are also of interest (Fig. 2). Through time there is a decrease in the differentiation between males and females affected by caries. Khok Phanom Di males and females show a marked difference in caries rates with females having almost three times the rate of males. This is a statistically significant result (FET = <0.0001). The Ban Lum Khao sample also shows a significant difference between male and female caries rates (FET = <0.0001). Again, the females have the higher rate, nearly five times that of the males. In the Iron Age sample from Noen U-Loke the sex difference is much less marked. Males from Noen U-Loke show a slightly higher caries rate (4.7 per cent) than females (3 per cent), a non-significant result (FET = 0.3). The number of subadult skeletons with permanent, fully erupted teeth in each sample is small. The mean age at death in each case is close to 10 years, but the comparability cannot be tested statistically. Age is clearly very important in subadults and interpretations must

25

20~b 15

LO Male
i Female

I 10

0 5i!01

,7
.^-

Khok Phanom DBan Lum Khao Noen U-Loke Figure2 Sex differencesin the correctedcariesrates (as listed in Table4) for all sites

76 N. Tayles,K. Domett and K. Nelsen Table6 Observedcariesratesin the subadults fromall sites,usingindividuals teeth as the units and of measurement
Observed caries (individual) n/total % Observed caries (tooth) n/total %

KhokPhanomDi Ban LumKhao Noen U-Loke

4/10 1/8 0/4

40.0 12.5 0

10/108 1/62 0/33

9.3 1.6 0

be made with caution. Of interest is that both the number of individuals and the number of teeth with caries (Table 6) show a decrease with time, a similar pattern to that seen in the results of the adult corrected caries rates. Nearly half of the ten subadults from Khok Phanom Di have one or more carious lesions (40 per cent). This percentage decreases by three times in the Ban Lum Khao sample of eight subadults (12.5 per cent), where only one subadult has a carious lesion. No caries are evident in any permanent teeth of the subadults from Noen U-Loke.

Discussion Our results show that there is no evidence in these three populations for an increase over time in caries prevalence, despite the increase in rice agriculture as a carbohydrate subsistence base. The data show the reverse pattern, with an overall decrease in caries over time. This is highly significant in females, and, although not significant in males, the prevalence is higher in the earliest site than in the later two. The subadult samples are very small, but are in accord with the adult evidence of a decrease in caries over time. There is also a decrease over time in the sex differences in caries rates. Females have more caries than males in the two earlier samples, but in the Noen U-Loke sample there is no difference in the rates between the sexes. This is the reverse of the expected pattern, where caries would be assumed to increase in females with agriculture. The relationship between diet and caries The development and progression of caries vary widely between individuals because of the complexity of caries aetiology. Although diet clearly plays a major role, there is not, unfortunately, a simple relationship. Numerous factors, both environmental and inherited, affect the development of caries. These include the genera of bacteria present in the mouth, morphology of the teeth, the composition of the saliva, the presence of minerals in the oral environment from diet and saliva, and food residues in the mouth. Age is also a factor in the development of caries for several reasons. First, the wear of the teeth can remove pits and fissures on the teeth which act as sites for caries formation, although a negative relationship between caries is not universal (Meiklejohn et al. 1992, cited in Lubell et al. 1994). Second, the dentine of the roots can be exposed to bacterial activity

Agriculture and dental caries? 77 through chronic periodontal disease or continued eruption of the teeth in response to wear of the occlusal surface. Third, age has an effect simply through the longer duration of the carious process (Thylstrup and Fejerskov 1994). However, despite the complexity of caries development on the individual level, the clinical literature nevertheless documents a clear correlation between diet, in particular the frequent consumption of carbohydrates (especially simple sugars), and a high level of dental caries at the population level. The process of the development of caries begins with the fermentation of carbohydrates. This occurs through the metabolism by the oral bacteria which results in an increase in acidity and production of proteolytic enzymes in the mouth. These attack the enamel, resulting in its demineralization. At first, this demineralization is reversible once the food residues are cleared and the oral pH increases. Demineralization will continue only if the pH does not increase, hence the frequency of eating carbohydrates and the subsequent length of time the pH is lowered will mediate the development of caries (Newbrun 1989; Thylstrup and Fejerskov 1994). The corollary of the frequency of eating is the 'clearance rate' - the speed at which oral pH rises again and food residues are removed from around the teeth. Related to this are the texture and degree of processing of the food, and therefore its availability for rapid fermentation. Another factor in the clearance rate is the salivary flow rate, which responds to the amount of chewing involved in eating the food (Thylstrup and Fejerskov 1994). Starch is the chief storage polysaccharide of plants. It is a polysaccharide of glucose, a large molecule in insoluble form, protected by a cell membrane which is only slowly attacked by enzymes in the saliva. Experiments with rats have shown that eating raw starch produces very little acid (Thylstrup and Fejerskov 1994). Processing and cooking partly breaks down the cell membrane, resulting in the faster fermentation of the residues and production of some acid (Lingstr6m et al. 1989). Sugar, by comparison, is a mono- or disaccharide, easily broken down and highly acidic and therefore highly cariogenic (Newbrun 1989). Carbohydrates in fresh fruit and vegetables, by contrast with other sugars, have a lower cariogenicity as they are eliminated rapidly, possibly within 5 minutes, because their texture requires chewing and therefore increases the salivary flow rate and the clearance rate. The consumption of soluble starch and sugar together is more cariogenic than sugar alone. The stickiness of starch reduces the clearance rate of the food and allows the effect of the sugar acids to be prolonged. It may also increase the acid production from the sugar (Lingstr6m et al. 1989, 1993). Most of the factors affecting caries development other than diet are unknown, and unknowable, for prehistoric people. Also, it is not possible to know what combinations of foods they ate, or how frequently they ate, but it is possible to assess the cariogenicity (or otherwise) of the carbohydrates and dietary sugars known to be available. Rice and caries - in prehistory and now As noted in the introduction, rice is documented as being present at Khok Phanom Di and Noen U-Loke. Although the evidence from the cemetery levels at Ban Lum Khao is not as yet clear, it was very unlikely that rice was not included in the diet. It is also most likely that the rice was eaten unpolished at all sites, as milling to produce the white rice preferred today is a recent development (FAO 1954). This would mean the rice was coarse

78 N. Tayles,K. Domett and K. Nelsen in texture, stimulating salivary flow and rapidly clearing the starchy material from the oral environment. The coarseness of the rice would indirectly affect caries development by its ability to wear the occlusal surfaces of the teeth, removing the pits and fissures which act as sites of plaque and bacteria accumulation. Currently, rice is central to all Southeast Asian diets, to the extent that the word 'rice' is synonymous with 'meal' or 'food' and 'to eat rice' with 'to eat'. All meals are rice-based, with relatively small amounts of accompanying fish, vegetables, eggs and sometimes meat (Hauck et al. 1959; Suvarnakich 1950; Wilson 1975). Whether this was a feature of prehistoric diet is an unknown. At least at Noen U-Loke, the amount of rice in the burials suggests it was a possibility. Other starchy foods in the prehistoric diet in Southeast Asia may have included starchy roots such as yams (Dioscorea spp) and Taro (Colocasea esculenta) as well as the possibility of the hypocotyls (germinating embryos) of one species of mangrove (Bruguiera sp.) (Thompson 1996), although the cariogenicity of these foods is unknown. Possible sources of sugars include bananas, which have been shown to be cariogenic (Mundorff-Shrestha et al. 1994) and 'palm sugar', made from the inflorescence of the coconut palm (Wilson 1975). There is no direct evidence of these other carbohydrate foods from any of the sites but the difficulty of obtaining botanical evidence is acknowledged. The assessment of the cariogenic potential of rice, as with other individual foods, is not easy because of the complexity of human diets. Experimental research is not ethical in humans because of the non-reversibility of caries. One rat-feeding experiment has shown that even 'Uncle Ben's' parboiled rice creates relatively small oral pH falls (Lingstr6m et al. 1993) compared with other carbohydrates. It appears that rice itself should have had relatively low cariogenicity, depending on how it was processed, what it was eaten with and how frequently. There is more evidence from epidemiological studies, although these are beset with difficulties of controlling other variables in the production of caries. Sreebny (1983) attempted to identify the role of cereals in caries on the population level, using the cereal supply compared with DMF (decayed, missing and filled) rates in 12-year-old children. He compared world-wide data on rice, wheat and maize supplies from countries with high and low consumption rates of each. He concluded that rice has no association with dental caries, maize has a negative correlation and wheat a positive correlation. However, this grand plan review is confounded by the complexity of caries aetiology and the association of levels of sugar consumption with socio-economic levels. The wealthiest countries eat wheat as their principal cereal and also have the highest sugar consumption levels. He suggests that at best cereals account for 20 per cent of caries prevalence compared with sugar which accounts for 50 per cent. Table 7 lists data on caries rates in modern populations with rice-based diets. This highlights some of the difficulties associated with these large-scale assessments. According to these data, the DMFT (decayed, missing, filled, teeth) rate in Thai 12-year-olds fell during the decade from the late 1970s to late 1980s. This seems unlikely when rates increased in Burma and Bangladesh and suggests some problems with the data. Nevertheless, the comparatively high rates for children in the cities of Bangkok and Chiang Mai compared with Khon Kaen and the village children in the north of Thailand could reflect two factors. The city children would have been eating more sugary foods and the rural children the

Agriculture and dental caries? 79 Table7 Rates of DMFT (decayed,missingand filled teeth) in modernchildrenin countrieswith rice-baseddiets
Country Date Source Age Cereals - % of diet Rice China 1950 1 12-27 Total 0.7 DMFT

Thailand Burma Thailand Bangladesh Burma Thailand Bangladesh Thailand, regional: ChiangMai Thailand, Bangkok? ChiangMai NorthThailand, ruralvillage KhonKaen Bangkok
Notes

1965 1970s 1970s 1970s 1989--90 1989-90 1989--90 1951 1953 1960 1960 -1995 -1995

2 3 3 3 4 4 4 5 6 7 7 8 8

10-14 12 12 12 12 12 12 14-16 7-14 9-16 9-16 12-16 12-16

76 71 73

78 73 85

0.6 0.8 2.7 1.8 1.5 1.1 3.5 1.5 1.9-2.0 1.97 0.27 2.35+/-2.22 4.74+/-2.97

fromLeatherwood al. (1965).3 Sreebny(1983).4 Woodward et 1 Afonsky(1951).2 Calculated and Walker(1994).5 Kridakara al. (1956).6 Kamalanathan al. (1960).7 Amatayukul(1960).8 et et et Kedjarune al. (1997).

so-called 'sticky' rice. Because of the method of processing, this is not completely decorticated and does not adhere to the teeth (Techanitiswad 1994). It also significantly reduces the drop in oral pH because it stimulates a high salivary flow rate (Kedjarune et al. 1997). These prevalence data are subject to the problems of veracity mentioned above, but show that rates were low in people eating rice-based diets in the controlled, early studies, increasing over time with 'development' (consistently accompanied by an increase in sugar in the diet), as in Bangkok and Chiang Mai. A large-scale study in Central China (Afonsky 1951), reported caries rates in children in boarding school on strictly controlled diets, where they had three meals a day, with no snacks. The diet was of crude polished rice and vegetables, with no sugar and very few other foods. The large sample of 3349 individuals aged 12-27 years had a very low mean DMFT rate of 0.7. In comparison, rates of caries in 'developed' countries during the same period are high. Large-scale reviews show that the highest mean rates in the late 1970s in ten out of fortyseven countries (all rated as 'developed' and with diets based on cereals other than rice) were 4.6-10.6 mean DMFT (Sreebny 1983) and in the late 1980s in fifteen out of 147 countries (again all 'developed') were 4.3-8.1 mean DMFT (Woodward and Walker 1994). Overall the evidence is that rice has a low intrinsic cariogenicity, particularly if it is crudely processed, so, despite the unknowns of frequency of eating, dental hygiene and other possible cariogenic constituents in the diet, there is no obvious reason for the development and intensification of rice agriculture to have increased caries rates in the populations affected.

80 N. Tayles,K. Domett and K. Nelsen Sex differences in caries rates As noted in the introduction, caries are often found to be more prevalent in females, in both prehistoric and modern populations (Larsen et al. 1991; Lukacs 1996; Legler and Menaker 1980). The reasons for this are not exactly known. Although tooth development and eruption are generally earlier in females than in males, this is in terms of months rather than years and therefore the longer exposure of the female teeth to potential decay is relatively minimal. The most probable explanation in prehistoric populations is a difference in diet, secondary to differences in the contribution of each sex to food acquisition. This is primarily based on the premise that the diet of females would be relatively high in carbohydrate and low in protein because they were more closely involved with the cultivation of the crops and had less access to protein from the animals hunted by the males (Larsen 1983; Larsen et al. 1991). A further assumption is that they also ate more frequently, with easy access to 'snack' foods around the home. As there is no opportunity for large-scale hunting by the rice-growing agriculturalists of Southeast Asia today, there is no ethnographic evidence for the same sexual division of labour and diet. We have observed that, currently in Thailand, both sexes work in the fields, although men tend to be responsible for fishing and fish farming. This contribution of both sexes to rice farming has been documented for subsistence farmers practising both swidden and irrigated farming in northwestern Thailand by Kunstadter (1985). In the dry season, both sexes engage in other activities such as pottery manufacturing or mat weaving, but the main occupation is farming (Mizuno 1978). Sex differences in caries rates in modern Thai are available as DMF rates from surveys carried out in the 1960s (Leatherwood et al. 1965). A sample of 890 females and 750 males aged from infancy to old age was drawn from all areas of Thailand. Almost all individuals were from rural areas, with a rice-based diet. It is assumed that the diet included some processed foods containing sugar. Females had minimally more caries than males from early childhood until early adulthood (mean DMFT at 20-24 years in females 0.8, males 1.2), but by middle adulthood the rates were the same in both sexes (mean DMFT 1.5). These data show that modern Thai DMF rates do not differ significantly between males and females, which is in accord with our evidence from the Iron Age site.

Does the adoption of agriculture necessarily lead to increased caries? The assumption of the positive association between caries and agriculture has become accepted to some extent in the palaeopathological literature. For example Larsen, in a review of biological responses to agriculture, states 'Dental caries clearly increases with agriculture' (1995: 188), and Lubell et al. say 'agricultural populations are expected to have uniformly high rates of pathology because of the cariogenic nature of carbohydrates' (1994: 211). To return again to Turner (1979), the author collated a large amount of data on mean caries rates in a wide variety of populations and from these argued that there was a universal increase in caries rates from a mean of 1.3 per cent (range 0.0-2.4 per cent; using teeth as the unit of measurement) in hunting and gathering societies to a mean of 10.4 per cent (range 2.1-26.9 per cent) in agricultural societies. However, there are deficiencies in this type of generalized comparison of data. First, the important problem of

Agriculture and dental caries? 81 the effect of inter-observer variation in the identification and recording of caries on the quality of the data is not acknowledged. Second, the 'agricultural' data included living populations, and some undated skeletal samples. The possible role of sugar in the diets of these people is not addressed, and, more importantly, the age effect is discounted, despite probable changes in longevity and certainly changes in the demographic profile of skeletal samples as a result of the increase in fertility and population size following the adoption of a sedentary lifestyle based on a reliable food source. The result is a drawing of a simple association between caries and agriculture. The acceptance of this concept has pervaded the literature to the extent that an increase in caries rates has even been used as evidence of the adoption of an agricultural lifestyle (Roosevelt 1984; Rose et al. 1984). This is despite the warning by Goodman et al. (1984) of circularity in this argument. 'If caries or attrition are used to help determine dietary changes, then they certainly can not be used to assess the impact of the change' (1984: 37).

Conclusions Although our data show a decrease in caries rates over time in both adults and children, and a reduction in sex differences in caries rates, we have clearly not presented definitive evidence of a negative correlation between the adoption of rice as a dietary base and caries, because of the complexity of caries aetiology, already discussed. We also have used a sample of only three sites. Khok Phanom Di is separated by up to 2000 years in time and 250km in distance from the Mun Valley sites. The environment was coastal compared with inland and riverine. There are possible genetic differences between the populations, the climate, geology and geography of the areas are different, and the components of the diet other than rice would have been different. The modern Thai data are consistent with our evidence, however, and we believe we have shown it is simplistic to make assumptions about such a significant event in prehistory as the adoption of an agricultural lifestyle solely on the basis of caries rates. As with most biological processes, the story is far more complex than can be explained by one variable alone. Department of Anatomy and StructuralBiology School of Medical Sciences University of Otago Dunedin

References on Afonsky,D. 1951.Some observations dentalcariesin centralChina.Journalof DentalResearch,
30: 53-61.

M. K., Amatayakul, Ladavalya, R. N. and Harris,R. 1960.A dentalsurveyof a smallgroupof Thai
children. American Journal of Clinical Nutrition, 8: 240-6. Cohen, M. N. and Armelagos, G. J. 1984. Paleopathology at the Origins ofAgriculture. Orlando, FL:

AcademicPress.

82

N. Tayles, K. Domett and K. Nelsen

FAO 1954. Rice and Rice Diets: A Nutritional Survey. FAO Nutritional Studies No. 1. Rome: Food and Agriculture Organization of the United Nations. Goodman, A. H., Martin, D. L., Armelagos, G. J. and Clark, G. 1984. Indications of stress from bone and teeth. In Paleopathology at the Origins of Agriculture (eds M. N. Cohen and G. J. Armelagos). Orlando, FL: Academic Press, pp. 13-50. Hauck, H. M., Hanks, J. R. and Sudsaneh, S. 1959. Food habits in a Siamese village. Journal of the American Dietetic Association, 35: 1143-8. Higham, C. F. W. and Bannanurag, R. 1990. The Excavation of Khok Phanom Di, a Prehistoric Site in Central Thailand, Vol. I, The Excavation, Chronology and Human Burials. London: Society of Antiquaries of London, Research Report XLVII. Higham, C. F. W. and Thosarat, R. 1998. Prehistoric Thailand: From Early Settlement to Sukhothai. Bangkok: River Books. Kamalanathan, G. S., Hauck, H. M. and Kittireja, C. 1960. Incidence of caries among school children in Bang Chan, Thailand, 1953. American Journal of Clinical Nutrition, 8: 231-46. Kedjarune, U., Migasena, P., Changbumrung, S., Pongpaew, P. and Tungtrongchitr, R. 1997. Flow rate and composition of whole saliva in children from rural and urban Thailand with different caries prevalence and dietary intake. Caries Research, 31: 148-54. Kridakara, O., Boozayaangool, R., Yuktananda, I. and Volker, J. F. 1956. Dental survey of selected Thai children: nutritional observations. American Journal of Clinical Nutrition, 4: 280-4. Kunstadter, P. 1985. Rice in a Lua' subsistence economy, northwestern Thailand. In Food Energy in Tropical Ecosystems (eds D. J. Cattle and K. H. Schwerin). New York: Gordon & Breach, pp. 21-44. Larsen, C. S. 1982. The anthropology of St. Catherines Island 3. Prehistoric human biological adaptation. Anthropological Papers of the American Museum of Natural History, 57(3). Larsen, C. S. 1983. Behavioural implications of temporal change in cariogenesis. Journal of Archaeological Science, 10: 1-8. Larsen, C. S. 1995. Biological changes in human populations with agriculture. Annual Reviews in Anthropology, 24: 185-213. Larsen, C. S. 1997. Bioarchaeology: Interpreting Behaviour from the Human Skeleton. Cambridge: Cambridge University Press. Larsen, C. S., Shavit, R. and Griffin, M. C. 1991. Dental caries evidence for dietary change: an archaeological context. In Advances in Dental Anthropology (eds M. A. Kelley and C. S. Larsen). New York: Wiley-Liss, pp. 179-202. Leatherwood, E. C., Burnett, G. W., Chandravejjsmarn,R. and Sirikaya, P. 1965. Dental caries and dental fluorosis in Thailand. American Journal of Public Health, 55: 1792-9. Legler, D. W. and Menaker, L. 1980. Definition, etiology, epidemiology and clinical implications of dental caries. In The Biologic Basis of Dental Caries (ed. L. Menaker). Hagerstown: Harper & Row, pp. 211-25. Lingstr6m, P., Holm, J., Birkhed, D. and Bj6rck, I. 1989. Effects of variously processed starch on pH of human dental plaque. Scandinavian Journal of Dental Research, 97: 392-400. Lingstr6m, P., Birkhed, D., Granfeldt, Y. and Bj6rck, I. 1993. pH measurements of human dental plaque after consumption of starchy foods using the microtouch and the sampling method. Caries Research, 27: 394-401. Lubell, D., Jackes, M., Schwarcz, H., Knyf, M. and Meiklejohn, C. 1994. The Mesolithic-Neolithic transition in Portugal: isotopic and dental evidence of diet. Journal of Archaeological Science, 21: 201-16. Lukacs, J. R. 1992. Dental paleopathology and agricultural intensification in South Asia: new evidence from Bronze Age Harappa. American Journal of Physical Anthropology, 87: 133-50.

Agriculture and dental caries?

83

Lukacs, J. R. 1995. The 'caries correction factor': a new method of calibrating dental caries rates to compensate for antemortem loss of teeth. International Journal of Osteoarchaeology, 5: 151-6. Lukacs, J. R. 1996. Sex differences in dental caries rates with the origin of agriculture in South Asia. CurrentAnthropology, 31: 147-53. Maloney, B. K. 1991. Palaeoenvironments of Khok Phanom Di: the pollen pteridophyte spore and microscopic charcoal record. In The Excavation of Khok Phanom Di. Vol. 2(1), The Biological Remains (eds C. F. W. Higham and R. Bannanurag). London: Society of Antiquaries of London, Research Report XLVIII, pp. 7-134. Meiklejohn, C., Wyman, J. M. and Schentag, C. T. 1992. Caries and attrition: dependent or independent variables? International Journal of Anthropology, 7: 57-62. Mizuno, K. 1978. The social organization of rice-growing villages. In Thailand: A Rice-Growing Society (ed. Y. Ishii). Honolulu: University Press of Hawaii, pp. 83-114. Molnar, S. 1971. Human tooth wear, tooth function and cultural variability. American Journal of Physical Anthropology, 34: 175-90. Molnar, S. and Molnar, I. 1985. Observations of dental diseases among prehistoric populations of Hungary. American Journal of Physical Anthropology, 67: 51-63. Mundorff-Shrestha, S. A., Featherstone, J. D. B., Eisenberg, A. D., Cowles, E., Curzon, M. E. J., Espeland, M. A. and Shields, C. P. 1994. Cariogenic potential of foods. Caries Research, 28: 106-15. Newbrun, E. 1989. Cariology. Chicago: Quintessence. Nikiforuk, G. 1985. Understanding Dental Caries 1. Etiology and Mechanisms: Basic and Clinical Aspects. Basel: Karger. Roosevelt, A. C. 1984. Population, health, and the evolution of subsistence: conclusions from the conference. In Paleopathology at the Origins of Agriculture (eds M. N. Cohen and G. J. Armelagos). Orlando, FL: Academic Press, pp. 559-84. Rose, J. C., Burnett, B. A., Nassaney, M. S. and Blaeuer, M. W. 1984 Paleopathology and the origins of maize agriculture in the Lower Mississippi Valley and Caddoan Culture areas. In Paleopathology at the Origins of Agriculture (eds M. N. Cohen and G. J. Armelagos). Orlando, FL: Academic Press, pp. 393-424. Sreebny, L. M. 1983. Cereal availability and dental caries. Community Dental and Oral Epidemiology, 11: 148-55. Suvarnakich, K. 1950. The Thai diet. Nutrition Reviews, 8: 289-91. Tayles, N. 1996 Tooth ablation in prehistoric Southeast Asia. International Journal of Osteoarchaeology, 6: 333-45. Techanitiswad, T. 1994. Determinants of the variation in dental caries of 2- and 4-year-old Thai children in an urban area of Khon Kaen Province, North Eastern Thailand. PhD thesis, University of Otago, Dunedin. Thompson, G. B. 1996. The Excavation of Khok Phanom Di, a Prehistoric Site in Central Thailand, Vol. 4, The Botanical Remains. London: Society of Antiquaries of London, Research Report LIII. Thylstrup, A. and Fejerskov, 0. 1994. Textbook of Clinical Cariology. Copenhagen: Munksgaard. Turner, C. G. 1979. Dental anthropological indications of agriculture among the Jomon people of Central Japan. American Journal of Physical Anthropology, 51: 619-36. Wilson, C. S. 1975. Rice, fish, and coconuts - the bases of Southeast Asian flavors. Food Technology, 6: 42-4. Woodward, M. and Walker, A. R. P. 1994. Sugar consumption and dental caries: evidence from 90 countries. British Dental Journal 23 April: 297-302.

Master your semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master your semester with Scribd & The New York Times

Cancel anytime.