The Arthritic Toll of Warfare, Farming, and Pastoralism among Populations of China and
Mongolia in the Zhou Dynasty/Xiongnu Period
Jacqueline T. Eng1, Vanchigdash Mergen2, Phillip L. Walker3, Shugang Yang4, Quanchao Zhang5, Hong Zhu5
1Dept.
Dept Anthropology,
Anthropology Western Michigan University
University, 2Dept.
Dept Archaeology & Anthropology
Anthropology, National University of Mongolia
Mongolia, 3Dept.
Dept Anthropology
Anthropology, UC
Santa Barbara, 4Henan Provincial Institute of Cultural Relics & Archaeology, 5Research Center for Chinese Frontier Archaeology, Jilin University
Abstract
Joints Observed
Statistical Results
Marked social changes occurred in East Asia during the first millennium BC, especially the
latter half. China experienced not only the Zhou dynasty’s fragmentation into contentious
hegemonic states, but also advances in agriculture and the introduction of iron technology.
In the northern steppe frontier, nomadic pastoral cultures arose and one of the strongest
confederacies to emerge were the Xiongnu, who posed a threat to Chinese states.
Bioarchaeological analysis of arthritic patterns in skeletal collections from sites in China
and Mongolia dating to this turbulent time offer insight into the toll of changing subsistence
strategies and increased warfare.
• Vertebrae: 1) OA = osteoarthritis of vertebral facets, 2) osteophytosis, or lipping of margins
• Shoulder: glenoid fossa of scapula, humeral head
• Elbow: distal humerus, proximal radius & ulna
• Wrist/hand (combined since limited hand bones): distal radius & ulna, carpals, hand bones
• Hip: acetabulum, femoral head
• Knee: distal femur, patella, proximal tibia
• Ankle/foot ((combined since limited foot bones):
) distal tibia & fibula,, tarsals,, foot bones
If an individual had any of the bones of a joint (with a complete joint surface), that person
was scored for presence of DJD (Fig. 2). Evidence of DJD on any joint surface (either left
or right) was sufficient to count the joint as affected. Recommendations for differential
diagnosis of DJD were followed, where eburnation (“polishing”) is a diagnostic criterion,
and in the absence of this, at least two of the following: osteophytes (joint margin new bone
growth), deformity of joint surface, porosity, and new bone growth on joint surface.8
Chi-square tests were used to analyze inter-site differences, while Fisher’s Exact tests
were used to determine which among the three had significantly different frequencies and
to compare frequencies between the sexes of each site.
Sites & Historical Context
China: An era of continual civil wars between hegemonic states punctuated the Eastern
Zhou period (1045-256 BC) of China, and warfare centered around the destruction of the
enemy.1 The state of Zhao (403-222 BC) was one of the last strong northern frontier states
(Fig 1a). The finds from Zhao-era sites in Helingeer County, in the modern province of
Inner Mongolia. derive from two cemeteries (“HT-HX”) of military soldiers (Fig. 1b). These
burials offer evidence of the rigors of military life during a time of intensified warfare that
shifted toward massed infantry, professional soldiers, and the use of cavalry pioneered by
123
the Zhao to deal with nomadic warriors.
warriors 1,2,3
A second site in central China offers a view of a non-military population from Henan
province, which served as a buffer zone between more powerful states during this time.2
The Chenjiagou cemetery (here referred to as “Henan,” Fig. 1b) practiced farming during a
time when innovations in iron technology and the use of oxen-draft made farming more
efficient.4 Their lifestyle contrasts with individuals from the other sites studied here who
were directly and continuously engaged in warring (Tab. 1).
Fig. 1. Maps of (a) Eastern Zhou states, note Zhao wall; (b) China sites; (c) Mongolia Xiongnu sites
Mongolia: Roughly contemporaneous to these two populations were horse-riding
horse riding cultures
that had spread eastward to the northern steppe frontier. Tensions between agrarian
Chinese states and nomads occupying the neighboring northern steppe became
increasingly evident as Chinese states erected defensive Great Walls along the north to
separate the states like Zhao from warlike nomads.5 A constant threat on these northern
borders were the Xiongnu confederation of nomadic tribes of the Ordos steppe region of
Mongolia, who eventually formed a nomadic empire (209 BC-AD 155). Burials of Xiongnu
samples analyzed here include those from multiple sites located throughout Mongolia
(Fig. 1c).
Tab. 1. Burial samples
Samples
Location
Subsistence
Henan
Central China Agriculture
Central China
HT‐HX
Inner Mongolia Agriculture
Xiongnu
Mongolia
Pastoralism
Occupation Total Adult
Farming
70
Soldier
133
Nom. Past.
73
Degenerative Joint Disease
Arthritis is a product of many factors, including degenerative changes to the joint from
functional (mechanical) stress, which can manifest skeletally as bone growth and/or bone
destruction. Despite debate on the etiology of specific causes of degenerative joint disease
(DJD), there is general agreement that joint use is an important contributor to
osteoarthritis.6 The distribution of arthritic changes in the skeleton is likely to vary between
societies and with different roles (e.g., gendered, specialist) within a society.7 Thus,
patterns of DJD can help reveal what, if any real differences exist among the populations
studied here, both between the groups and within their groups.
Fig. 2. DJD seen in samples
Fig. 3. Distribution of sexes
among the sites
Results
Results were obtained with respect to the frequencies of joint disease within each age
category. As expected there were generally higher frequencies among the older age
groups. Here we limit the discussion to the comparison of results from the total population
across the sites (Fig. 5; Tab. 2).
Tab. 2. DJD rates across sites by sex and total population
Wrist/
Hand
Hip
Knee
Ankle/
Foot
Obs. M
3/21
10/20
6/19
11/26
7/23
4/20
7/19
2/13
%M
Vert.
Vert.OA osteophy Shoulder Elbow
14%
50%
32%
42%
30%
20%
37%
15%
Obs. F
1/19
6/19
3/18
13/24
1/16
0/17
10/22
1/16
%F
17%
32%
17%
54%
6%
0%
46%
6%
9/37
24 / 50
%Total
Xiongnu
8/39
4/37
17/41
3/29
41%
24%
48%
21%
11%
42%
10%
5/5
19/42
22/47
17/26
6/60
27/67
7/29
100%
45%
47%
65%
10%
40%
24%
Obs. F
0/2
2/2
9/19
7/21
2/10
2/39
17/40
1/13
%F
0%
100%
47%
33%
20%
5%
43%
8%
Obs. Tot
2/7
7/7
34/70
30/78
19/42
9/106
52/119
15/51
%Total
Obs. M
10%
16/39
2/5
%M
Fig. 5. Total rates of DJD by site and joint
4/40
40%
Obs. M
Total Female Between F samples
Chi-sq p-value
(Fisher's Exact)
0.435
0.805
4.24
0.120
5.54
0.063
7.01
0.030
Henan > Xiongnu
1.15
0.564
5.19
0.075
8.04
0.018
Both > Xiongnu
1.55
0.461
-
Discussion & Conclusion
Fig. 4. Age and sex distribution
of samples in Henan (right) and
Xiongnu (left).
Obs Tot
Obs.
Between M samples
Total Male
Chi-sq p-value
(Fisher's Exact)
1.81
0.405
4.46
0.108
1.13
0.567
1.04
0.593
6.2
0.045
HT-HX > Henan
2.67
0.263
0.791
0.673
0.469
0.791
Within-site comparisons of M vs. F: only three significant instances, all where male rates
were greater than that of females within their group: wrist/hand for Xiongnu (p=0.04)
and HT-HX (p=0.02), and the ankle/foot for Xiongnu (p=0.04).
Age: A second potential source of bias is in the age composition within each sample, as
joint disease is often slowly progressive,6 with individuals of advanced years more likely to
show arthritic changes. Samples were examined with respect to three broad age
categories of Young Adults (YA=16-30), Middle Adults (MA=30-45), and Older Adults
(OA=45+), as well as a category of adults for whom an age category could not be assigned
(A). The HT-HX sample was comprised of many commingled remains, most lacking
associated pelves and crania, so many were not aged into specific categories and instead
placed in the general “Adult” category, though most were likely in the young to middle adult
ages since this was a military population. Also, the few that did have their ages determined
were young adults. There are some differences in age categories present for the Xiongnu
and Henan samples (Fig. 4).
Henan
TOTAL Population Between Tot samples
Chi-sq
p-value
(Fisher's Exact)
1.80
0.407
9.00
0.011
HT-HX > both
6.61
0.037
HT-HX > Henan
4.78
0.092
Henan > Xiongnu
5.68
0.058
HT-HX > Henan
5.00
0.082
Xiongnu > HT-HX
7.34
0.025
HT-HX > Xiongnu
7.00
0.030
HT-HX > Xiongnu
Inter-site comparisons of the sexes (Tab. 3):
• Males: difference only in the wrist/hand joint, where the HT-HX rate was greater than
that of Henan (p=0.02).
• Females: Henan had a higher rate of elbow DJD than did Xiongnu (p=0.01), while both
Henan (p=0.02) and HT-HX (p=0.01) had greater rates of knee DJD than Xiongnu.
Potential Biases: Sex & Age
HT-HX
Specifically, as tested in this study, these differences may be correlated with:
•Different modes of subsistence: agriculture in China vs. nomadic pastoralism in Mongolia
•Different occupations & activities: mounted warriors vs. soldiers of China vs. farmers
•Sexual division of labor within each population
Tab. 3. Statistical results (significant χ2 results in bold and Fisher’s Exact results noted)
Vert. OA
Vert. Osteophy.
Shoulder
Elbow
Wrist/hand
Hip
Knee
Ankle/foot
Sex: Males outnumber females in all three sites, though the
sex difference between the sites is not significant (χ2=1.3,
p=0.52, Fig. 3). The over-representation of males is marked
in the Xiongnu sample with 63% males and 34% females
females,
close to the level of significance (χ2=3.2, p=0.07). Similarly
there are more males (56% M vs. 34% F, with the rest of
indeterminant sex) in HT-HX, also close to the level of
significance (χ2=3.6, p=0.06). The sex ratio is also not
significantly different in the Henan sample (56% M, 44% F,
χ2 =0.5, p=0.50).
From: http://www.chinahighlights.com/greatwall/history/
Inter-site total population: HT-HX often had significantly higher frequencies of joint disease,
but the pattern differed with respect to the joint location and with the comparative
population (Tab. 3). For vertebral osteophytosis, HT
HT-HX
HX had a greater frequency than the
other sites (p=0.01,Henan; p=0.004, Xiongnu). HT-HX had higher rates than that of Henan
in DJD of the shoulder (p=0.02) and wrist/hand (p=0.02). In the lower limb, HT-HX had
greater rates of DJD in the knee (p=0.01) and ankle/foot (p=0.03) compared to Xiongnu,
which did have higher rates of DJD in the hip compared to HT-HX (p=0.04).
29%
100%
49%
39%
45%
9%
44%
29%
6/34
17/33
17/38
14/39
14/32
9/43
12/38
7/36
%M
18%
52%
45%
36%
44%
21%
32%
19%
Obs. F
2/21
6/21
5/25
4/23
3/20
4/22
2/21
0/21
%F
10%
29%
20%
17%
15%
18%
10%
0%
Obs. Tot
8/56
23/55
22/64
18/64
17/54
13/65
14/60
7/58
%Total
14%
42%
34%
28%
32%
20%
23%
12%
During this turbulent period, the soldier sample from HT-HX seems to have suffered the
most joint disease, likely related to occupational activities such as repeated drills and
training, as well as actual combat. Compared to another “active warrior” population of the
nomadic pastoral Xiongnu, the HT-HX group suffered from more back problems, as well as
knee and foot joint disease, while the Xiongnu sample had more hip disease, which may
be expected since the Xiongnu were mounted warriors while the HT-HX sample may have
been composed of mostly foot soldiers. Though not at the level of significance, Xiongnu
females also had higher rates of hip DJD than the females of the other two sample
populations, possibly indicative of their participation in horse riding.
The agricultural sample from Henan had the greatest elbow DJD rate, particularly among
women. DJD related to functional use in the limb joints is most often correlated with the
elbow, and secondly the knee.9 The Henan elbow DJD rate may be linked to their farming
activities in this age of agricultural intensification. However, compared to the other sites,
this sample had the least amount of DJD in the back, shoulder, wrist/hand, and ankle/foot,
which suggests that overall, the other two populations engaged in more strenuous activity
in the course of military life and/or nomadic pastoralism. The prevalence of middle aged
adults in the Henan sample may explain their lower DJD rates, but the HT-HX sample, an
active soldier group, had the overall highest rates and likely comprised younger adults.
Furthermore, the Xiongnu sample still had high rates of DJD despite its higher
representation
p
of yyounger
g age
g g
groups.
p
Within each population, gendered activities seem possible as males generally had higher
rates of DJD than females in most joints except a few instances. These results may
represent the more strenuous, active life of males as farmers, soldiers, and mounted
warriors. Future studies will benefit from a more thorough exploration of patterns of DJD
within each group for sidedness and polyarticular DJD, as well as studies in the association
between occupation and traumatic lesions.
Acknowledgements:
We would like to thank funding sources that supported travels to Asia (UCSB ISBER grant to P. Walker, and to
J. Eng the MHC travel award, and WMU’s SFSA and IEFDF awards), as well as the institutions that housed
collections and facilitated observations (Henan Provincial Institute of Cultural Relics and Archaeology, Jilin
University, and the National University of Mongolia). Much thanks to Andrew Baker for help with data and
photo sorting, and Jsaon Glatz for generating the maps. We thank Phil for inspiring us all.
References: 1. Yates, R.D.S., Early China, in War and Society in the Ancient and Medieval Worlds. Asia, The Mediterranean, Europe, and Mesoamerica., K.
Raaflaub and N. Rosenstein, Editors. 1999, Harvard University Press: Cambridge. p. 7-45. 2. Lewis, M.E., Warring States political history, in The Cambridge History
of Ancient China: From the Origins of Civilization to 221 B.C., M. Loewe and E.L. Shaughnessy, Editors. 1999, Cambridge University Press: New York. p. 587-650. 3.
Underhill, A.P., Warfare and the development of states in China, in The Archaeology of Warfare, E.N. Arkush and M.W. Allen, Editors. 2006, University Press of
Florida: Gainesville. p. 253-285. 4. Hsu, C.-Y., The Spring and Autumn Period, in The Cambridge History of Ancient China: From the Origins of Civilization to 221
B.C., M.E. Lewis and E.L. Shaughnessy, Editors. 1999, Cambridge University Press: New York. p. 545-586. 5. Di Cosmo, N., The northern frontier in pre-imperial
China, in The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., M. Loewe and E.L. Shaughnessy, Editors. 1999, Cambridge University
Press: Cambridge. p. 885-989. 6. Ortner, D.J., Identification of Pathological Conditions in Human Skeletal Remains. 2nd ed. 2003, New York: Academic Press. 7.
Ortner, D.J., Description and classification of degenerative bone changes in the distal joint surfaces of the humerus. American Journal of Physical Anthropology,
1968. 28: p. 139-155. 8. Rogers, J. and T. Waldron, A Field Guide to Joint Disease in Archaeology. 1995, New York: John Wiley & Sons. 9. Jurmain, R.D.,
Degenerative changes in peripheral joints as indicators of mechanical stress: occupational limitations. International Journal of Osteoarchaeology, 1991. 1: p. 247-252.
Contact: jacqueline.eng@wmich.edu