Paget's Disease in an Archeological Population


  • The authors have no conflict of interest


The aim of this investigation was to study the prevalence and distribution of Paget's disease in an archeological population. Paget's disease, first described over 100 years ago, is a well-recognized chronic disorder involving abnormal bone turnover with established radiological features. Prevalence within modern populations varies both within individual countries and between them. Paleopathological examples are uncommon and sporadically reported both from Europe and the Americas and from many periods of history. A large skeletal assemblage (2770 individuals) from Barton on Humber, UK, provided an opportunity to examine the prevalence of Paget's disease in one area of the northern England over the period 900-1850 AD. All bones were examined visually for evidence of Paget's disease of the bone (PDB) and all abnormal bones were examined further by plain radiography. Fifteen cases of probable Paget's disease were found. The overall prevalence was 2.1% in those aged >40 years. The prevalence before 1500 AD was 1.7% and post-1500 AD was 3.1%. The distribution of disease mirrored modern disease, with the lumbar spine, pelvis, and proximal femur being the commonest sites. The prevalence of Paget's disease in the United Kingdom over the last 1000 years has been assessed. Although there is a trend of increasing prevalence, this did not reach statistical significance. This is likely caused by the small sample size, but this is by far the greatest number of cases of PDB described in a single skeletal assemblage to date. The distribution of lesions within the skeleton is unchanged.


Paget's disease of bone (PDB) or osteitis deformans, which was first described and named after Sir James Paget in 1877, is a chronic disease involving disruption of the mechanism of normal bone turnover. It can be characterized as a localized disorder of bone remodeling with an increase in osteoclastic mediated bone resorption and a compensatory increase in new bone formation. This results in a disorganized structure of woven and lamellar bone and, consequently, bone enlargement. The nature of the architecture of Pagetic bone predisposes it to fragility. Vascularity is increased also. The majority of cases is asymptomatic and is discovered incidentally by radiography or raised serum alkaline phosphatase. Bone pain may occur and joint pain and stiffness may result if the involved bone is adjacent to a joint. Symptoms due to complications of the disease include bowing deformities, fractures, cranial nerve compression, high output heart failure, and the development of primary bone tumors such as osteosarcoma. Pagetic bone has a pathognomonic tile-like or mosaic pattern at histological examination. The radiological appearances of PDB are characteristic and depend on the stage of the disease. Typically, an initial lytic phase commences in the subarticular region of a bone and extends distally, classically with a flame-shaped leading edge. In subsequent phases, the development of new bone results in, first, a mixed and, finally, a sclerotic appearance. Eventually, the bone becomes expanded with a coarse trabecular pattern and bowing of long bones is a common feature.

The etiology of PDB is unknown. Various theories have been suggested, although no definitive proof exists for any etiological agent to date. Any proposed etiology must account for the striking variation in geographical prevalence of PDB both throughout the world and within individual countries. Longitudinal studies of how PDB has changed over time both in prevalence and in skeletal distribution provide valuable information that may help to confirm or refute etiological theories. The earliest reported case of PDB in archeological human remains was in the 20s,(1) in which a femur from a French Neolithic site displayed the characteristic radiological changes. Wells and Woodhouse(2) later reported an Anglo-Saxon example from Durham. Medieval examples have been reported in Norwich(3) and a 16th century case was reported in Wells.(4)

This study examines the prevalence of PDB in a large population of skeletons from the South Humberside region deriving from the 10th to the 19th centuries. A comparison is made between the prevalence during the period 1000-1500 and that during 1500-1900. The etiology of PDB in view of the findings of this study is discussed.


The skeletons available for study were derived from an excavation at St. Peter's Church, Barton on Humber, UK, during the late 1970s and early 1980s (Fig. 1). Nearly 3000 skeletons were recovered, approximately one-third being children. Inclusion criteria for entry into the study were that the skeleton should be aged over 35 years (by standard anthropological criteria)(5) and with at least 40% of skeletal parts present. Six hundred sixty-seven skeletons fulfilled the entry criteria. All bones were examined visually for signs of PDB. Signs included increased size, thickness, or vascularity. All suspected bones were examined further by plain radiography, as were all other bones of any affected skeletons. Because all bones with an abnormal appearance or shape were radiographed, two further cases of PDB were recognized from the assemblage. One case was discovered during the radiological examination of fractures and another case was found during a study of the appearance and prevalence of hyperostosis frontalis interna.

Figure FIG. 1..

Map of the United Kingdom showing location of Barton on Humber and Lancashire towns with high prevalence rates.


Fifteen cases of PDB were diagnosed at Barton, UK. About 40% of the cases were suspected on visual examination. The Paget's cases came from all phases of the site with seven cases from before 1500 and eight cases after 1500. The male/female ratio was 11:3, which is a little higher than modern populations(6) (Table 1).

Table Table 1.. PDB Cases Identified at Barton on Humber
original image

By definition, all the skeletons identified with PDB at Barton had at least 40% of skeletal parts present and all were assessed as being over 35 years old. The prevalence for the population as a whole was 2.2%. With a varying number of individuals fulfilling the selection criteria within the different phases, it was decided to assess the prevalence in larger numbers by studying greater time intervals. Thus, the population was divided into two groups: pre-1500 and post-1500. Two individuals were in phases straddling the divide—one was placed in the pre- and the other in the post-1500 group. The denominator population likewise was divided equally. There were seven cases of PDB in the 409 skeletons fulfilling the selection criteria in the pre-1500 group, a prevalence of 1.7%. In the post-1500 group, there were eight cases of PDB in 258 skeletons, a prevalence of 3.1% (Table 2). Examples of Pagetic bones from several skeletons are shown, together with corresponding plain radiographs (Fig. 2,3,4 and 5)

Table Table 2.. Prevalence of PDB at Barton on Humber
original image
Figure FIG. 2..

(A) Distal femur of adult female dating from 18th-19th century showing features of PDB (bone expansion and coarse trabeculation). (B) Corresponding plain radiograph of distal femur of adult female dating from the 18th-19th century showing features of PDB (bone expansion and coarse trabeculation).

Figure FIG. 3..

(A) Lumbar vertebrae of adult male dating from the 10th-13th century showing features of PDB (bone expansion and coarse trabeculation). (B) Proximal femora of the same adult male dating from the 10th-13th century showing features of PDB on the right (bone expansion and coarse trabeculation). (C) Corresponding plain radiograph of lumbar vertebrae of adult male dating from the 10th-13th century showing features of PDB (cortical thickening and coarse trabeculation).

Figure FIG. 4..

(A) Phalanx of adult male dating from the 12th-13th century showing features of PDB (bone expansion and trabecular coarsening). (B) Corresponding plain radiograph of phalanx of adult male dating from the 12th-13th century showing features of PDB (cortical thickening and coarse trabeculation). (C) Normal phalanx of same adult male for comparison. (D) Corresponding plain radiograph of normal phalanx of same adult male for comparison.

Figure FIG. 5..

(A) Os calcis of adult male dating from the 18th-19th century showing features of PDB (trabecular coarsening). (B) Corresponding plain radiograph of os calcis of adult male dating from the 18th-19th century showing features of PDB (cortical thickening and coarse trabeculation).


This study examines the prevalence of PDB in this part of the United Kingdom over the past 1000 years. The difference in prevalence between the pre-1500 and post-1500 populations shows a clear trend of increasing prevalence although not statistically significant by the κ2 test. The small number of cases of PDB found in the assemblage makes statistical significance difficult to achieve. However, this is by far the greatest number of cases of PDB described in a single skeletal assemblage to date. Therefore, the trend of increasing prevalence is an important finding, while not reaching levels of statistical significance.

A possible source of bias in these results reflects the difficulty of accurately establishing those skeletons >45 years of age at death. It is possible that there were more elderly individuals in the post-1500 population group because of increasing life expectancy. This may partly explain the higher prevalence in the later population. However, there is no means of testing this hypothesis with currently available anthropological aging methods.

There was a significant amount of disarticulated skeletal material recovered around the graves during the excavation. These bones were not included in the study because they did not fulfill the entry criteria of at least 40% of skeletal parts being present. However, some of this material was examined also and a definite identification of at least three Pagetic bones was made. Two of these were sacra and one was a femoral shaft. They displayed sufficient abnormality on visual inspection to be selected for radiographic examination. Very few of the disarticulated bones were so examined. Thus, further unidentified cases of PDB may exist among this material. One of the sacra, from context 83, was examined microscopically, confirming the diagnosis.(4) The disarticulated bones were derived from locations sufficiently different from each other and the Pagetic skeletons discussed previously to make it unlikely that they were associated.

The distribution of affected parts does not seem to differ from that recognized today. The pelvis, sacrum, and lumbar vertebrae are the most frequently involved bones followed by the proximal femur. This pattern of anatomical distribution is very similar to that described in large studies carried out during the last 100 years(7,8) (Table 3). It is interesting that 2 out of the 15 cases showed PDB in the phalanges.(7) According to a large study of 889 patients with PDB, the phalanges are involved in <2% of cases. This may relate to the fact that the hand constitutes only 2% of the total skeletal volume. However, the true frequency of phalangeal involvement may be underestimated in the literature. Because the majority of patients with PDB are asymptomatic, the diagnosis usually is an incidental finding and the phalanges are less commonly imaged than the axial skeleton. In addition, the features may be subtle and more difficult to detect than in large bones.

Table Table 3.. Anatomical Distribution of PDB
original image

The etiology of PDB is unknown. The main areas of research have concentrated on the relative contribution of genetic and environmental factors. It has been shown that ∼12% of affected patients have a first-degree relative with PDB,(9) supporting a genetic basis for the etiology. However, this does not exclude environmental factors. Several studies have established linkage to a marker on chromosome 18(10) and linkage to human leukocyte antigen (HLA) D antigens.(11) It has been suggested that there are two distinct types of PDB, familial and sporadic, which can be distinguished on their clinical features and may have different etiologies.(9,12)

Numerous possible environmental agents may be involved in the etiology of PDB. Low levels of calcium intake in childhood as measured by milk consumption have been shown to be associated with an increased risk of developing PDB.(9) An infectious agent is an attractive possibility that may partly explain localized clusters of high prevalence. Evidence in favor of a viral etiology is conflicting. Viral-like inclusion bodies containing paramyxovirus antigens have been identified by electron microscopy in Pagetic osteoclasts.(13,14) However, these inclusion bodies are not only found in Pagetic osteoclasts, but also in osteoclasts from diseases known not to have a viral origin such as pyknodysostosis(15) and oxalosis.(16) Therefore, their discovery in Pagetic osteoclasts may be purely incidental. Immunohistochemical studies have shown positive staining for a variety of viral antigens in Pagetic bone, including the measles virus.(17) In situ hybridization studies using reverse-transcription polymerase chain reaction have shown both measles and canine distemper virus in tissue from patients with PDB.(18) However, other studies have failed to replicate these findings.(19) Serological tests do not show increased titers of antibodies to canine distemper or measles virus in patients with PDB.(20,21) Circumstantial evidence in favor of the canine distemper virus as an etiological agent in PDB has been proposed in studies showing an association between past dog ownership and PDB(22,23) and between ownership of dogs unvaccinated for canine distemper and PDB.(24) However, other studies have failed to replicate these findings.(25–27) Thus, the evidence for a viral etiology in PDB is far from conclusive.

Therefore, to date, no consensus exists on genetic versus environmental etiological factors in PDB. Studies of both geographical variation and time trends in prevalence provide valuable data that may help distinguish between genetic and environmental influences.

One of the earliest accurate studies of geographic variation in prevalence of PDB was carried out in 14 towns in the United Kingdom in 1977.(28) They found overall prevalence rates of 5.4% in people aged >55 years with a focus of high prevalence in three tightly clustered Lancashire towns, Preston, Bolton, and Blackburn (Fig. 1). A follow-up study in 1980(29) confirmed this high prevalence in three other nearby towns in the Lancashire region. Prevalence rates throughout the United Kingdom varied widely from 8.3% in Lancaster to 2.3% in Aberdeen, but no obvious reason was identified for this. A further study of prevalence in 13 towns throughout Europe(30) showed wide variation in prevalence. No country was found to have prevalence as high as that in the United Kingdom. Highest rates were found in French towns (2.7-2.0%), intermediate rates in German towns (1.1-1.3%), and lowest rates in Scandinavian towns (0.4-0.5%). Other studies have shown low prevalence rates of PDB in Ireland (0.7-1.7%)(31) and Norway.(32) Reports of prevalence of PDB in Africa suggest that the disease is uncommon,(33) although there may be localized areas of higher prevalence.(34) It is possible that the prevalence of PDB in Africa is underestimated because of the lower average age of patients in hospitals in comparison to those in Europe and the relatively poorer access to radiological facilities. However, studies of prevalence of PDB in American blacks have shown that the rates are not significantly different from the white population.(35) Similar findings have been reported among the south African black population.(36) This lends support to an environmental etiology.

Few studies have focused on possible time trends in the prevalence of PDB. A follow-up of the 1977 UK prevalence study found a dramatic decline in the prevalence of PDB in 10 British towns between 1977 and 1999.(37) Indirect measures of prevalence of PDB have been carried out by examination of rates of mortality attributed to Paget's and to osteosarcoma. These figures suggest a decline in prevalence of PDB between 1870 and 1915.(38) The severity of PDB may be declining in New Zealand.(39) Patients are significantly older at diagnosis and have less severe disease as measured by serum alkaline phosphatase.

However, it is tempting to suggest that changes in prevalence of PDB that have occurred during the last millennium may be attributed to the spread of an environmental agent through a genetically susceptible population. The rapidity of the recent decline in prevalence in the United Kingdom particularly favors an environmental rather than genetic etiology.

Therefore, current evidence points to a combination of a genetic predisposition and a superimposed environmental factor(s) in the etiology of PDB. This study supports that hypothesis. However, it is difficult to compare the prevalence rates of this study with those in the literature. Many of the quoted studies assess prevalence in patients over 55 years of age, whereas this study measures prevalence in the over 35-year age group. This is because of the difficulty in accurately aging skeletons of an older age. It is tempting to propose that an environmental factor, most prevalent in the United Kingdom, was responsible for gradual changes in the prevalence of PDB over the last millennium. However, during the last 50 years a steep decline in prevalence of the disease has occurred in the United Kingdom, perhaps because of reduced exposure to or improved treatment of an environmental agent.


This study was funded by the Arthritis and Research Campaign and the English Heritage.