Content uploaded by Meltem Çetin
Author content
All content in this area was uploaded by Meltem Çetin on Mar 04, 2020
Content may be subject to copyright.
ORIGINAL ARTICLE
Serpil Savas áMeltem CË etin áMehmet Akdog
Æan
Nurettin Heybeli
Endemic ¯uorosis in Turkish patients: relationship
with knee osteoarthritis
Received: 16 April 2001 / Accepted: 18 July 2001 / Published online: 10 August 2001
ÓSpringer-Verlag 2001
Abstract Fluoride excess primarily eects dental and
skeletal tissues, leading to a condition known as endemic
¯uorosis. The radiological and clinical features of en-
demic ¯uorosis vary in dierent parts of the world. The
aim of this study was to investigate the clinical and
radiological features of endemic ¯uorosis in Turkish
patients. Physical examination and radiological investi-
gations were performed in 56 patients with endemic
¯uorosis and 40 age- and sex-matched controls. Knee
osteoarthritis (OA) was the main abnormality in both
groups, both clinically and radiologically. The radio-
logical severity of knee OA was greater in the endemic
¯uorosis group than in controls (P=0.01). Osteophytes
at the tibial condyles and superior margin of the patellar
articular surface of the femur, polyp-like osteophytes on
the non-weight-bearing medial side of the femoral con-
dyle, and popliteal loose bodies were detected more
frequently in the endemic ¯uorosis group than in con-
trols (P=0.0001). We suggest that the presence of
atypically located osteophytes in the knees may be a
feature of endemic ¯uorosis in Turkish patients and that
endemic ¯uorosis may increase the severity of OA in the
knees.
Keywords Bone áCartilage áFluoride áEndemic
¯uorosis áKnee joint
Introduction
Endemic skeletal ¯uorosis is a chronic metabolic bone
and joint disease caused by ingestion of large amounts of
¯uoride either through water or rarely from foods of
endemic areas [1]. Although the prevalence of this dis-
ease has decreased considerably, it still occurs in some
parts of the world such as Isparta, an endemic ¯uorosis
area in southern Turkey. A wide range of radiological
features have been described in endemic skeletal ¯uo-
rosis [2, 3, 4, 5, 6, 7, 8]. The main radiological appear-
ances are osteosclerosis, osteopenia, membranous and
ligamentous calci®cation, exostoses, coarse trabecular
pattern, diaphyseal widening, and intermittent growth
lines. Joint pains, some crippling deformities such as
limited spinal movement, thoracic kyphosis and ¯exion
deformities at the hips and knees, bony leg deformities
(genu varum, genu valgum, bowing, wind-swept), and
spinal cord compression due to the vertebral osteoscle-
rosis have been reported as the clinical ®ndings [4, 5, 6,
7, 9, 10, 11, 12, 13, 14].
To the best of our knowledge, endemic skeletal ¯u-
orosis has not been reported previously from Turkey.
Observing a high number of female patients with dental
¯uorosis complaining of knee pain in Isparta led us to
investigate the clinical and radiological features of these
patients.
Patients and methods
Fifty-six patients with endemic ¯uorosis who attended the Su
Èley-
man Demirel University physical therapy and rehabilitation out-
patient clinic for various painful conditions were included in the
study. The clinical diagnosis of endemic ¯uorosis was modi®ed
from the criteria of Wang et al. [2]: (1) living in the endemic ¯uo-
rosis region since birth, (2) having mottled tooth enamel, indicating
dental ¯uorosis, (3) consuming water with ¯uoride levels above
Rheumatol Int (2001) 21: 30±35
DOI 10.1007/s002960100132
S. Savas (&)
Su
Èleyman Demirel University Medical School,
Physical Therapy and Rehabilitation Department,
Isparta, Turkey
M. CË etin
Su
Èleyman Demirel University Medical School,
Radiology Department, Isparta, Turkey
M. Akdog
Æan
Su
Èleyman Demirel University Medical School,
Biochemistry Department, Isparta, Turkey
N. Heybeli
Su
Èleyman Demirel University Medical School,
Orthopaedics and Traumatology Department, Isparta, Turkey
Posta Kutusu 76, 32000, Isparta, Turkey
1.2 ppm (normal 1 ppm), and (4) a urine ¯uoride level greater than
1.5 mg/l (normal <1.5 mg/l).
Forty patients living in a nonendemic region who attended the
same outpatient clinic and had nontraumatic knee pain were ran-
domly selected as controls. The patients and controls were all ur-
ban dwellers and housewives. The medical histories and the main
complaint of the groups were recorded. Patients with metabolic
bone disease or in¯ammatory disease in their medical histories were
excluded. Physical examination was performed in all groups. The
patients' body weight and height were recorded. Body mass index
(BMI) was calculated for each patient as weight (kg)/height (m)
2
.
Diagnosis of genu varum or valgum deformity was made using the
femorotibial angle. This angle is formed by the intersection at the
knee joint of the long axes of the femur and tibia, represented by
appropriate lines drawn on the radiograph. Measurements of 175°,
or 5°valgus, were accepted as normal [15].
Biochemical investigations
Serum samples were analyzed for calcium, inorganic phosphorus,
total alkaline phosphatase, total proteins, albumin, creatinine, and
blood urea nitrogen by standard biochemical methods. Patients
and controls were asked to bring tap water from their houses in
plastic containers. Water, serum, and urine samples were estimated
by using an ion-speci®c electrode (Orion F 94±09).
Radiology
The patients with endemic ¯uorosis had anteroposterior (AP) ra-
diographs made of the forearm, hands, pelvis with the upper end of
the femur, thoracic spine, lumbar spine, knees (anteroposterior, or
AP, and lateral views), and tibia and ®bula (AP and lateral views).
Controls had AP and lateral views of the tibia and ®bula only. All
radiographs were analyzed by two radiologists who were blinded to
each other's ®ndings.
Radiological changes were classi®ed as: (1) existence of radio-
logical knee abnormalities based on the worse knee consisting of
femorotibial osteoarthritis (OA) graded according to the Kellgren
method using a 5-point scale (0±4), osteophyte formation at the
superior and inferior margins of patella, osteophytes on the ante-
rior and posterior parts of the tibial condyles and at the superior
margin of the patellar articular surface of the femur, and popliteal
loose bodies, (2) osteosclerosis, an overall increase in bone density
with thickening of the cortex, and (3) ossi®cation of ligaments,
tendons, and interosseous membranes. De®nite OA was de®ned as
that with a Kellgren score of 2 [16]. Maximum patellar osteophyte
size (grade 3) was determined by measuring with a ruler. Grades 1±
2 were calculated as two-thirds and one-third reductions in the area
of grade 3 [17].
Statistical analysis
Results were given as meanSD. Statistical signi®cance was set at
the 0.05 level. Comparison of the groups was performed using
Student's-ttest in parametric values and the chi-squared test in
nonparametric values. Correlation analysis was performed to assess
the relation between some radiological and the clinical features.
Results
Demographic and clinical features
The endemic ¯uorosis group consisted of 51 female and
®ve male patients and the control group included 34
females and six males. There was no dierence in sex
distribution between the groups (P=0.36). The mean
age of the endemic ¯uorosis group and controls was
54.911.6 years and 57.459.76 years, respectively
(P=0.26). The BMI of both groups were 29.755.08
and 29.364.31, respectively. (P=0.69). The mean
numbers of years since menopause were 12.16.3 in
endemic ¯uorosis patients and 11.356.7 in controls.
There was no dierence between the groups in the
number of mean years since menopause (P=0.5).
The most common complaint of the patients with
endemic ¯uorosis was knee pain (37 patients, 66.1%).
The distribution of complaints of the patients with
endemic ¯uorosis is shown in Table 1.
Table 1 The prevalence of symptoms in the 56 patients with
endemic ¯uorosis
Location N
Knee 37 (66.1%)
Hip 2 (3.6%)
Feet 2 (3.6%)
Low back 3 (5.4%)
Low back and knee 5 (8.9%)
Generalised pain 4 (7.1%)
Shoulder 2 (3.6%)
Neck 1(1.8%)
Fig. 1 Radiograph of knee showing osteophytes on the anterior
and posterior parts of the tibial condyles (thin arrows) and at the
superior margin of the patellar articular surface of the femur
31
None of the patients had thoracal kyphosis, ®xed
knee, or hip ¯exion deformities. Seventeen patients
(30.4%) with endemic ¯uorosis and ®ve controls
(12.5%) had genu varum deformity (P=0.05). The mean
femorotibial angle was 184.63.9°in the right knee and
184.44.2°in the left knee in the endemic ¯uorosis
patients with genu varum deformity. In controls with
genu varum deformity, the mean femorotibial angle of
the right knee was 183.62.6°, and the mean femorot-
ibial angle of the left knee was 183.81.3°. Genu varum
deformity correlated to age (P=0.0001, r=0.48) but not
to BMI (P=0.32, r=0.13) in our patients with endemic
¯uorosis. In controls, genu varum deformity correlated
to BMI (P=0.01, r=0.38) but not to age (P=0.07,
r=0.28). None of the patients in either group had genu
valgum deformity.
Biochemical parameters and ¯uoride levels
Serum total calcium, inorganic phosphorus, and alkaline
phosphatase concentrations were normal in all patients.
The mean ¯uoride level in the drinking water was
signi®cantly higher in patients with ¯uorosis
(2.740.85 mg/l) than in controls (0.710.06 mg/l)
(P<0.0001). Serum and urine ¯uoride levels were
0.100.04 mg/l and 1.960.37 mg/l, respectively, in the
patients with endemic ¯uorosis. These values were sig-
ni®cantly higher than in controls, whose mean serum
and urine ¯uoride levels were 0.030.01 mg/l and
0.370.16 mg/l, respectively (P=0.0001).
Radiology
The main radiological features of the patients with en-
demic ¯uorosis were degenerative changes including
mainly the medial femorotibial and patellofemoral
compartments. Twelve patients (21.4%) with endemic
¯uorosis and eight control patients (20%) had grade 2
OA, 16 patients (28.6%) with endemic ¯uorosis and
three control patients (7.5%) had grade 3 OA, and six
patients (10.7%) with endemic ¯uorosis and one control
patient (2.5%) had grade 4 OA. The OA severity was
greater in the endemic ¯uorosis group (P<0.01).
Osteophytes at the anterior and posterior parts of the
tibial condyles, osteophytes at the superior margin of the
patellar articular surface of the femur (Fig. 1), and
popliteal loose bodies (Fig. 2) were detected more fre-
quently in the endemic ¯uorosis group than in controls.
Fig. 2 Popliteal loose bodies
Fig. 3 Large osteophytes located at the superior and inferior parts
of the patella
32
There was no dierence between groups in the existence
of osteophytes at superior and inferior margins of the
patella (P=0.67). However, these osteophytes were
larger in endemic ¯uorosis patients than in controls
(P=0.001) (Fig. 3).
Twenty-®ve (44.6%) patients with endemic ¯uorosis
had polyp-like osteophytes at the medial non-weight-
bearing margin of the femoral condyle (Fig. 4). This
kind of osteophyte was not detected in controls. Com-
parison of the radiological features of the groups are
summarized in Table 2.
Two patients (3.6%) with endemic ¯uorosis had axial
osteosclerosis (Fig. 5), and ®ve (8.9%) had interosseous
membrane calci®cation on the forearm (Fig. 6). None of
the patients with endemic ¯uorosis had cartilage calci-
®cation or bowing of the tibia or ®bula.
Discussion
In our patients with endemic ¯uorosis, the knee joint
was the most commonly involved joint, both radiologi-
cally and clinically. In skeletal ¯uorosis, involvement of
the axial skeleton is characteristic, and changes are most
marked in the spine, pelvis, and forearm [5, 6, 10, 18,
19]. Knee involvement in endemic ¯uorosis has been
clinically described mainly in children and adolescents as
¯exion deformities [6, 19] and genu varum or valgum
deformities due to the bowing of long bones caused by
osteomalacia, secondary hyperparathyroidism, and
rickets [4, 9, 10, 11]. Not much is known about the ra-
diological features of the knee involvement in endemic
¯uorosis.
Table 2 Comparison of knee radiographs of the patients with endemic ¯uorosis and controls
Fluorosis Controls P
Grades 2 and 3 osteophytes at the superior and inferior margins of the patella 21 (37.5%) 3 (7.5%) 0.001
Osteophyte at the anterior part of the tibial condyle 29 (51.8%) 3 (7.5%) 0.000
Osteophyte at the posterior part of the tibial condyle 27 (48.2%) 2 (5%) 0.001
Osteophyte at the superior margin of the patellar articular surface of the femur 25 (44.6%) 8 (20%) 0.013
Popliteal loose bodies 20 (35.7%) 4 (10%) 0.004
Fig. 4 Polyp-like osteophytes on the medial non-weight-bearing
margin of femoral condyle Fig. 5 Osteosclerosis of the spine
33
It was surprising to see female dominance in an en-
demic region. As OA is particularly common in middle-
aged/elderly females in this geographic region, it is
possible that femorotibial OA was the main problem in
both groups; however, the severity of OA was greater in
patients with endemic ¯uorosis than controls, both
clinically and radiologically. We thought that endemic
¯uorosis might be responsible for the increased severity
of degenerative features. Some of the osteophytes seen in
endemic ¯uorosis patients were located in unusual sites
for OA such as the tibial condyles. Some of the osteo-
phytes were polyp-like and located at the non-weight-
bearing medial side of the medial femoral condyle,
which is atypical for OA. The osteophytes at the supe-
rior and inferior margins of the patella were larger than
the patellar osteophytes seen in controls. These atypical
osteophytes were not seen in the patients with endemic
¯uorosis without knee pain, so we decided that they were
not incidental ®ndings.
Fluoride is a cumulative poison that increases meta-
bolic turnover of the bone in favor of bone formation
[1]. It stimulates bone cell proliferation by directly in-
hibiting osteoblastic acid phosphatase activity [20] and
by prolonging or enhancing the mitogenic signals of
growth factors [21, 22]. Histopathological studies of
bone in ¯uorosis have shown osteoid tissue deposited
irregularly on the trabeculae and cortex, with extension
into muscle attachments [6]. This osteoblastic activity
causes a marked increase in bone formation at the organ
level, causing the production of exostoses, calci®cation
of tendons and ligaments, and osteosclerosis [22]. The
other explanation for the degenerative changes in
the knees of our patients with endemic ¯uorosis may
be the chondrotoxic eect of ¯uoride. Fluoride mainly
aects bone metabolism [21]. However, cartilage is one
of its deposition areas [22, 23]. In an animal study [9], it
was shown that excessive ¯uoride ingestion caused
necrosis of articular chondrocytes, ulcer formation, and
articular calci®cation, which causes a mushroom-like
appearance.
Radiological changes in industrial ¯uorosis suggest
that physical strain on bones, ligaments, and joints plays
an important role in the development of the lesions [24,
25]. The reason for the selective knee involvement in our
patients may be ¯uoride's selectivity for the most stres-
sed joints [26, 27]. Sitting habits like hyper¯exion sitting
and ritual worship in our population may stress the
knees and cause vulnerability to ¯uorotic damage.
In our study, some radiological ®ndings such as
osteosclerosis, interosseous membrane calci®cation, or
ligament calci®cation, which were accepted as hallmarks
of skeletal ¯uorosis, were not found as frequently as in
the literature [1, 10, 22]. We also did not observe severe
crippling deformities. The causes for the clinical and
radiological dierences may be due to several factors
such as the nature, dose, duration of ¯uoride exposure,
age, sex, dietary habits, or their combination [1, 3, 6, 10,
18]. The mean ¯uoride level in the drinking water of our
region was 2.7 ppm. Osteosclerosis and interosseous
membrane calci®cation have usually been reported in
levels over 4 ppm [6, 10]. The total quantity of ingested
¯uoride is reported as the single most important factor
determining the clinical course of this disease, which is
characterized by immobilization of joints of the axial
skeleton and of major joints of the extremities [1, 6].
Jolly [10] suggested that crippling deformities are usually
associated with drinking water with a 10 ppm level for
10 to 20 years. However, the concentration of ¯uoride
alone was not found to be responsible for the incidence
of skeletal ¯uorosis. In dierent areas with similar ¯u-
oride content in the drinking water [10, 28] and even in
the same endemic area where the source of drinking
water was the same, the incidence of the disease and
extent of bone changes diered, and X-ray ®ndings of
bone ¯uorosis were signi®cantly dierent [7, 10]. Arnala
suggests the variability of individual susceptibility to
¯uoride [29]. Another important cause is nutritional
status. It is accepted that poor nutrition and low calcium
intake enhance the deleterious eects of ¯uoride [3, 4,
14, 18]. As we did not assess the nutritional status of the
patients, we cannot clearly state that poor nutrition is
responsible for the abnormal radiological ®ndings in our
patients with endemic ¯uorosis.
We suggest that atypically located osteophytes in the
knees may be a feature of endemic skeletal ¯uorosis in
Fig. 6 Ossi®cation of interosseous membrane
34
Turkish patients. We also suggest that endemic ¯uorosis
may increase the severity of knee OA. Further studies
are needed in order to understand the exact mechanism
of bone and cartilage changes in the knees of patients
with endemic ¯uorosis.
Acknowledgement We are grateful to Dr. Tayfun Turgut for his
expert opinion on X-rays.
References
1. Krishnamachari KA (1986) Skeletal ¯uorosis in humans: a re-
view of recent progress in the understanding of the disease. Prog
Food Nutr Sci 10:3±4, 279±314
2. Wang Y, Yuming Y, Gilula LA, Wilson AJ (1994) Endemic
¯uorosis of the skeleton: radiographic features in 127 patients.
AJR 162:93±98
3. Mithal A, Trivedi N, Gupta SK, Kumar S, Gupta RK (1993)
Radiological spectrum of endemic ¯uorosis: relationship with
calcium intake. Skeletal Radiol 22:257±261
4. Jackson WPU (1962) Further observations on the Kenhardt
bone disease and its relation to ¯uorosis. S Afr Med J 10:932±
936
5. Siddique AH (1955) Fluorosis in Nalgonda district, Hydera-
bad-Deccan. BMJ 10:1408±413
6. Singh A, Jolly SS, Bansal BC, Mathur CC (1963) Endemic
¯uorosis. Epidemiological, clinical and biochemical study of
chronic ¯uorine intoxication in Panjab (India). Medicine
42:229±246
7. Jolly SS, Singh BM, Mathur OC (1969) Endemic ¯uorosis in
Punjab (India). Am J Med 47:553±563
8. Haimanot RT, Fekadu A, Bushra B (1987) Endemic ¯uorosis in
the Ethiopian Rift Valley. Trop Geogr Med 39:209±217
9. Jun-chao X, Yun-zhao W, Dian-min X et al (1987) X-ray
®ndings and pathological basis of bone ¯uorosis. Chin Med J
(Engl) 100:8±16
10. Jolly SS, Singh BM, Mathur OC, Malhotra KC (1968) Epide-
miological, clinical and biochemical study of endemic dental
and skeletal ¯uorosis in Punjab. BMJ 16:427±429
11. Christie DP (1980) The spectrum of radiological bone changes
in children with ¯uorosis. Radiology 136:85
12. Haimanot RT (1990) Neurological complications of endemic
skeletal ¯uorosis, with special emphasis on radiculo-myelopa-
thy. Paraplegia 28:244±251
13. Maloo JC, Radhakrishnan K, Thacker AK, Mousa ME (1990)
Fluorotic radiculomyelopathy in a Libyan male. Clin Neurol
Neurosurg 92:63±65
14. Teotia M, Teotia SP, Singh KP (1998) Endemic chronic ¯uoride
toxicity and dietary calcium de®ciency interaction syndromes of
metabolic bone disease and deformities in India: year 2000.
Indian J Pediatr 65:371±381
15. Bauer GCH, Insall J, Koshino T (1969) Tibial osteotomy in
gonarthrosis (osteoarthritis of the knee) Am J Bone Joint Surg
51:1545
16. Kellgren JH, Lawrence JS (1963) Atlas of standard radiographs
of arthritis, vol 2: the epidemiology of chronic rheumatism.
Blackwell, Oxford
17. Nagaosa Y, Mateus M, Hassan Lanyon P, Doherty M (2000)
Development of a logically devised line drawing atlas for
grading of knee osteoarthritis. Ann Rheum Dis 59:587±595
18. Fisher RL, Medcalf TW, Henderson MC (1969). Endemic ¯u-
orosis with spinal cord compression. A case report and review.
Arch Intern Med 149:697±700
19. Teotia M, Teotia SPS, Kunwar KB (1971) Endemic skeletal
¯uorosis. Arch Dis Child 46:686±691
20. Lau KHW, Farley JR, Freeman TK, Baylink DJ (1989) A
proposed mechanism of the mitogenic action of ¯uoride on
bone cells: Inhibition of the activity of an osteoblastic acid
phosphatase. Metabolism 38:858
21. Gruber HE, Baylink DJ (1991) The eects of ¯uoride on bone.
Clin Orthop 267:264±277
22. Boivin G, Chavassieux P, Chapuy MC, Baud CA, Meunier PJ
(1989) Skeletal ¯uorosis: Histomorphometric analysis of bone
changes and bone ¯uoride content in 29 patients. Bone 10:89±
99
23. Bang S, Boivin G, Gerster JC, Baud CA (1985) Distribution of
¯uoride in calci®ed cartilage of a ¯uoride-treated osteoporotic
patient. Bone 6:207±210
24. Czerwinsky E, Nowak J, Dabrowska D, Skolarczyk A, Bartl-
omiej K (1988) Bone and joint pathology in ¯uoride exposed
workers. Arch Environ Health 43:340±343
25. Boillat M, Garcia J, Velebit L (1980) Radiological criteria of
industrial ¯uorosis. Skeletal Radiol 5:161±165
26. Farley SM, Libanati CR, Schulz EE et al (1984) Fluoride
therapy in osteoporosis had rapid eects on both the axial and
peripheral skeleton which can be detected by increased serum
alkaline phosphatase. Calcif Tissue Int 36:A41
27. Schulz EE, Libanati CR, Farley SM, Kirk GA, Baylink DJ
(1984) Skeletal scintigraphic changes in osteoporosis treated
with sodium ¯uoride: Concise communication. J Nucl Med
25:651
28. Kaminsky LS, Mahoney MC, Leach J, Melius J, Miller MJ
(1990) Fluoride: bene®ts and risks of exposure. Crit Rev Oral
Biol Med 1:261±281
29. Arnala I, Alhava EM, Kauranen P (1985) Eects of ¯uoride on
bone in Finland. Acta Orthop Scand 56:161±166
35