close

Вход

Забыли?

вход по аккаунту

?

The longitudinal course of hand osteoarthritis in a male population.

код для вставкиСкачать
1323
THE LONGITUDINAL COURSE OF HAND
OSTEOARTHRITIS IN A MALE POPULATION
DOUGLAS A. KALLMAN, FREDRICK M. WIGLEY, WILLIAM W. SCOTT, JR.,
MARC C. HOCHBERG, and JORDAN D. TOBIN
In this prospective analysis of the natural history
of osteoarthritis (OA) of the hand, the incidence and
progression of various radiographic features of OA were
examined in 177 men who participated in the Baltimore
Longitudinal Study of Aging. Subjects who were less
than 60 years old were selected based on the availability
of at least 4 hand radiographs taken during at least 20
years of followup, while subjects who were age 60 or
older were required to have had at least 4 hand radiographs taken during only 14 years of followup. Individual joints of the hands were graded for the presence and
severity of 5 features of OA: osteophytes, joint space
narrowing, subchondral sclerosis, lateral deformity,
and cortical collapse. They were also graded according
to the global KellgredLawrence scale. The incidence
and rate of progression of OA, as measured by all scales,
increased in progressively older age groups. Cox’s proFrom the Applied Physiology Section, Gerontology Research Center, National Institute on Aging; the Department of
Medicine, Rheumatology Division at Francis Scott Key Medical
Center, and the Department of Radiology, Skeletal Radiology Section, Johns Hopkins University School of Medicine; and the Department of Epidemiology, Johns Hopkins University School of
Hygiene and Public Health, Baltimore, Maryland.
Douglas A. Kallman, MD: Gerontology Research Center,
National Institute on Aging; Fredrick M. Wigley, MD: Department
of Medicine, Rheumatology Division at Francis Scott Key Medical
Center, Johns Hopkins University School of Medicine; William W.
Scott, Jr., MD: Department of Radiology, Johns Hopkins University School of Medicine; Marc C. Hochberg, MD, MPH: Department of Medicine, Johns Hopkins University School of Medicine,
and Department of Epidemiology, Johns Hopkins University School
of Hygiene and Public Health; Jordan D. Tobin, MD: Gerontology
Research Center, National Institute on Aging.
Address reprint requests to Jordan D. Tobin, MD, Gerontology Research Center, 4940 Eastern Avenue. Baltimore, MD
21224.
Submitted for publication June 6, 1989; accepted in revised
form March 20, 1990.
Arthritis and Rheumatism, Vol. 33, No. 9 (September 1990)
portional hazards models showed that isolated doubtful
osteophytes (Kellgren/Lawrence grade 1) and isolated
joint space narrowing predicted the development of the
radiographic features of OA. The rate of OA progression slowed as the severity increased. We conclude that,
in men, the incidence of radiographic features of hand
OA increases with age. Regardless of age, the earliest
radiographic signs of OA are joint space narrowing and
doubtful osteophytes.
The natural history of osteoarthritis (OA) of the
hand remains uncertain. Although there have been a
number of cross-sectional studies of hand OA (1-71,
very little is known about its longitudinal progression
(8.9). The prevalence of hand OA increases with age
(1-7). but it is not known whether older subjects are at
increased risk of developing hand OA, i.e., whether
there is a higher incidence of hand OA in the elderly.
To date, there have been no studies estimating the
incidence of hand OA. In addition, the earliest radiographic signs of OA are unknown.
Since cartilage degeneration is commonly considered to be the initial pathologic defect in OA (lo), it
is reasonable to expect that cartilage loss, reflected by
radiographic joint space narrowing (1 1,12), would be
an early sign of OA. Once OA is present, however, it
is not known how rapidly it progresses or whether the
rate of progression is constant throughout the patient’s
life span and throughout the course of the disease.
Without longitudinal studies, it is impossible to address these questions.
Radiographic changes, as defined by Kellgren
and Lawrence (13), have been accepted by the World
Health Organization since 1961 (14) as the method of
choice for diagnosing and classifying OA in epidemio-
1324
KALLMAN ET AL
logic studies. The KellgredLawrence scale is based on
the presence of osteophytes, which is considered the
pathognomonic and earliest feature of OA. However,
many joints do not easily fit into this grading system.
Hand joints may be narrowed and sclerotic (probable
signs of OA), without the presence of osteophytes.
Although there is some evidence from animal models
that the osteophyte is the first sign of OA (15,16), several
studies of humans have questioned the importance of
osteophytes alone (without other radiographic evidence
of OA), in the knee (17-19) and hip (20).
Recently, we developed grading scales for the
individual radiographic features of hand OA: osteophytes, joint space narrowing, subchondral sclerosis,
lateral deformity, and cortical collapse (21). In the
present study, we applied this method of grading hand
radiographs to men who were studied prospectively
for 14-29 years. We report the prevalence and incidence of each radiographic feature of OA, rates of
progression, and risk factors for the development and
progression of hand OA. This study provides insight
into the natural history of hand OA.
SUBJECTS AND METHODS
Subjects and radiographs. Subjects participating in
the Baltimore Longitudinal Study of Aging routinely undergo posteroanterior radiography of the left hand, without
intensifying screens or magnification, at 2-4-year intervals.
These apparently healthy men drawn from the community
represent a highly educated, middle-to-upper socioeconomic
group who are participating in a study of aging and whose
characteristics have been described previously (22). For this
study, 177 men were selected based on their length of
followup, without consideration of hand OA status. Subjects
were required to have had at least 4 hand radiographs
performed and to have been followed for at least 20 years
(mean 2 SEM 23.5 ? 0.25 years) if they were younger than
age 60 at the time of entry into the study and for at least 14
years (mean 2 SEM 16.9 0.45 years) if they were age 60 or
older at the time of entry into the study. All subjects were
right-handed.
Radiographic scoring. On each hand radiograph, individual joints were graded for the presence and severity of
5 selected individual features of OA: osteophytes, joint
space narrowing, subchondral sclerosis, lateral deformity,
and cortical collapse (21). Osteophytes and joint space
narrowing were differentiated into 4 ordinal grades (grades 0,
1, 2, and 3), while sclerosis, lateral deformity, cortical
collapse, and narrowing of the first carpometacarpal (CMC I )
joint were scored as either absent or present.
Joints were also graded using the KellgrenlLawrence
scale (13,14). Grade 2 of the KellgrenlLawrence scale corresponds to grade 1 of the osteophyte scale (definite small
osteophyte). Grade 1 of the KellgrenLawrence scale (osteo-
phytes of doubtful significance) corresponds to grade 0 of the
osteophyte scale (none).
Reading radiographs. Radiographs were assigned
random numbers, and all identifying marks were obscured.
All radiographs were read individually in random sequence
and were not paired. Readings were performed over a
3-week period by 1 experienced reader (DAK); interreader
reliability of DAK’s interpretations compared with those of
others (FMW, WWS, and MCH) and intrareader reliability
over time had previously been established (21). A subset of
the radiographs read for this study had also been read 8
months earlier by the same reader, and the intrareader
reliability on this subset was good to excellent for all scales
(intraclass correlation coefficient 0.66-0.93) (23).
The 5 distal interphalangeal (DIP) and 4 proximal
interphalangeal (PIP) joints, the CMCl joint, and the trapezioscaphoid (TS) joint were each scored individually, using
both the individual features scales and the Kellgrenl
Lawrence scale. The CMCl joint and the TS joint were not
scored for cortical collapse, and the TS joint was not scored
for osteophytes and lateral deformity, because these findings
are not characteristic of OA in these joints. Kellgren and
Lawrence did not develop a method for scoring the TS joint.
Data analysis. Subjects were arbitrarily divided into 3
age groups prior to data analysis: young (<40 years; mean 2
SEM 34.9 2 0.5, n = 46), middle-aged (4&59 years; mean ?
SEM 48.3 2 0.5, n = 85), and old (260 years; mean ? SEM
65.6 2 0.6, n = 46). The prevalence of positive findings per
100 joints was calculated for each joint using each grading
scale.
For the purpose of these analyses, the Kellgrenl
Lawrence scale was considered positive if grade 1 or greater
changes were noted. The chi-square test for trend was used
to test the significance of increasing prevalence in progressively older age groups (24). Incidence rates were calculated
for each feature by defining incident cases as those initially
negative subjects who developed a positive finding during
the study period. The time until conversion was designated
as the time at risk for each subject. If a subject did not
convert, then the entire period of followup was considered
time at risk. Incidence rates are reported as new cases per
1,000 person-years at risk.
Longitudinal analysis of OA development and progression was performed with life tables (24) and Cox’s
proportional hazards regression models (25). The most severely affected DIP and PIP joints (“highest DIP” and
“highest PIP”) were used as variables for longitudinal
analysis. The progression of OA in the CMCl and TS joints
was also determined. The progression of OA by 1 grade
signified an event, and observations were right-censored if
there was no OA progression during the followup period. A
subject was excluded from longitudinal analysis if that
subject had a maximum score at the start of the study, and
the OA could therefore not progress.
For the life-table analysis, differences in the rate of
progression between the age groups were tested with rank
statistics for homogeneity using the log rank test, which
places more weight on longer survival times, and the
Wilcoxon signed rank test, which places more weight on
shorter survival times (26). The effect of age and starting
level of each of the scales on the development and progres-
1325
COURSE OF HAND OA IN MALES
UMO
YFCO
K' 0 0 37
0.0 I 7
"13 38 59
S ' O
0 4
D O 0 2
c o o 2
DIP 5
DIP 4
YMO
K 7 617
0'0 1 7
N' 2 18 26
YMO
K 2 1 9
0 2 0 2
"13 24 35
K' 2 26 30
0' 0 2 1.1
N' 4 25 52
S ' O
0 7
D
O
0
0
c
o
o
2
K'11 26 39
0' 2 4 20
N' 7 22 39
s'o 0 7
X-MLL
C'O
0
4
DIP 2
DIP 3
K' 4 21 48
0' 0 5 13
N' 4 21 48
S ^ O 0
4
~~
K'
2
6 15
N" 2 17 17
YMO
K' 0 1 22
0 ' 0
0 7
N' 0 9 1 1
o
o
o
s o o o
s
o
o
o
D ' O
0
4
D
D
O
0
0
s
O
0
2
c o o o
c o o o
c o o o
PIP 5
PIP 4
PIP 3
c o o o
PIP 2
DIP 1
YLO
K'25 43 67
0' 2 8 31
N'O
0 7
s o 0 4
D O 0 2
c - - -
K' 7 1 2 30
0' 2 1 15
"13 27 4 1
D ' O
0
N' 0
s o
5 11
0 7
CMC 1
7
c o o o
TS
Eiahest P I P
Figure 1. The prevalence of positive findings of hand osteoarthntis (OA), per 100 subjects at the time
of study entry, for each radiographicfeature (percent of subjects with score of 1 or more) in each joint.
Subjects were divided into 3 age groups: young (Y;<40 years), middle-aged (M:40-59 years), and old
(0;260 years). OA was measured by the Kellgren/Lawrence scale (K) and by the presence of
osteophytes (O), joint space narrowing (N),subchondral sclerosis (S), lateral deformity (D), and
cortical collapse (C). Highest DIP = most severely involved distal interphalangealjoint; Highest PIP
= most severely involved proximal interphalangealjoint; CMC 1 = first carpometacarpal joint; TS =
trapezioscaphoid joint. The chi-square test for trend was used to measure the significance of increasing
prevalence of OA with age. * = P < 0.05; t = P < 0.01; t = P < 0.001.
sion of OA was determined using Cox's proportional hazards
regression model (27,28). Cox's models quantify the effect
and statistical significance of each risk factor (covariate).
The relative risk (RR) of OA progression was calculated as
the exponent of the regression coefficient (ecoefficient
) of each
risk factor; 95% confidence intervals (CI) were calculated as
e(coefficient +
The following Cox's models were tested: 1) models including age and individual risk factors
considered one at a time, with dummy variables used to
account for different grades of the risk factor, and 2) stepwise multivariate models, with an entry level set to include
all variables, to adjust for possible collinearity of independent risk factors.
RESULTS
Prevalence. The percentage of subjects with
positive findings of OA of the hand (grade 1 or more) at
entry into the study is shown in Figure 1, for each of
the features in each joint. The prevalence of each
finding in the highest DIP and highest PIP joints, each
considered as a group, is also shown. There was a
significant trend toward higher prevalence rates of
hand OA in older age groups as measured by the
Kellgren/Lawrence scale and by osteophytes, joint
space narrowing, and subchondral sclerosis in the DIP
and CMCl joints. Each radiographic feature of OA
was most prevalent in the DIP joints, although definite
osteophytes were most common in the CMCl joint.
The PIP and TS joints were less commonly involved.
Osteophytes (including doubtful osteophytes, Kellgredlawrence grade 1) and joint space narrowing
were the most common radiographic findings in the
DIP joints. Subchondral sclerosis was less common
1326
KALLMAN ET AL
Middle
Aged
(JQ(-J-jJ
u
SUEROSIS
Old
LJ
SUERUSIS
LJ
SCLEROSIS
Figure 2. The prevalence of positive findings (grade 1 or higher) of
hand osteoarthritis in the distal interphalangeal joints, per 100
subjects, divided by age group (young <40 years, middle-aged 40-59
years, and old 860 years). A, Interrelationship of definite osteophytes (grade 1 of the osteophyte scale, grade 2 of the Kellgrenl
Lawrence scale), joint space narrowing, and subchondral sclerosis.
B, Interrelationship of osteophytes (grade I of the Kellgred
Lawrence scale), joint space narrowing, and subchondral sclerosis.
space narrowing became increasingly more common in
older age groups (Figure 2B). Joint space narrowing
was a much more common isolated finding than was
definite osteophytes (Figure 2A). The combination of
joint space narrowing and definite osteophytes also
increased with age. Sclerosis was seen most commonly in conjunction with both joint space narrowing
and osteophytes, and was never seen in the absence of
joint space narrowing (Figures 2A and B).
Incidence. Figure 3 shows the incidence rates
(per 1,000 person-years at risk) for each age group for
the highest DIP joint only. The incidence of each joint
finding of OA was highest in the DIP joints. New cases
of OA were less common in the PIP, CMC1, and TS
joints (data not shown). In general, the incidence of
each of the features increased with age. Cox’s proportional hazards models were used to test the significance of the increased risk of incident cases in older
subjects, using age at the start of the study as a risk
factor. For each feature in each joint group, except for
PIP joint space narrowing and doubtful osteophytes of
the CMCl joint, the risk of developing OA increased
significantly with age. Thus, not only did older subjects have a higher prevalence of OA (Figure 2), but
those older subjects who did not have OA were still
more likely to develop new disease.
Risk factors for osteoarthritis development. Osteophytes and joint space narrowing were the only
1201
YOUNG
MIDDLEAGED
than definite osteophytes (osteophyte scale grade 1,
KellgredLawrence scale grade 2), which were less
common than joint space narrowing. Lateral deformity
and cortical collapse were uncommon findings, which
were never observed in young or middle-aged subjects. Joint space narrowing was the most common
finding in the PIP joints.
Cross-sectional relationships between features.
The interrelationships between the common individual
features of OA in the DIPjoints (vide supra) are shown
in Figures 2A and B, using Venn diagrams for each age
group. In young subjects, isolated doubtful osteophytes (KellgredLawrence grade 1) or joint space
narrowing were the most common findings (Figure
2B).
the older age group, isolated doubtful osteophytes continued to be a more common finding,
isolated joint space narrowing became less common.
The combination of doubtful osteophytes and joint
rn om
New Cases
per Thousand
Person-Years
at Risk
KELLOAEN
OSTEO
-NO
SCLEROSIS
0ER)AMIN
COLLAPSE
SCALE
Figure 3. The incidence of hand osteoarthritis (number of new
cases per 1,OOO person-years at risk) as determined by each radiographic feature. The significance of increasing incidence with age
was tested using Cox’s proportional hazards regression models and
was significant at P < 0.01 for the KellgrenlLawrence scale and at P
< 0.001 for the remaining scales. Osteo = osteophytes. See Figure
1 for other definitions.
1327
COURSE OF HAND OA IN MALES
Table 1. Cox's proportional hazards regression models for the development of osteoarthritis*
Outcome
variable
Risk
factor
Osteophyte
(n = 143)"
Age
Narrowing
Kellgren
(n = 85)
Age
Narrowing
Grade
NO. of
subjects
at risk
NO. of subjects with
change
0
1
2
3
91
45
7
0
48
38
6
0
1.05
1
1.94
2.45
-
I .03-1.08
1.26-2.98
1.03-5.82
1.01-1.07
0.75-2.21
0.5CL28.26
-
0
1
2
3
Narrowing
(n = 97)
Age
Osteophytes
Age
Kellgren
Sclerosis
(n = 163)
Age
Narrowing
0
I
2
3
0
1
2
3
4
0
I
Osteophytes
2
3
0
1
2
Age
Narrowing
Kellgren
0
1
2
3
0
I
2
3
4
64
20
I
0
91
5
-
RR
-
1.04
56
19
I
0
1
1.29
3.75
-
I .06
1
0.91
10.39
-
95% c1
1
1
0
0
-
I .04-1.09
0.3 2-2.60
1.04-103.63
1.02-1.08
I . 19-3.56
0.45-3.97
1.63-176.65
-
97
57
9
0
143
16
4
-
I .07
1.04-1.10
22
33
9
0
46
14
4
1
1
0
-
64
27
5
97
57
9
0
85
58
16
4
0
69
4
1
0
43
26
4
-
-
1.05
1
2.06
1.33
16.97
2.61
22.15
1
2.73
7.54
1.05
22
1
33
9
0
14
32
14
4
0
2.30
14.95
-
I
2.99
6.03
16.12
-
1.494.56
9.1 1-53.86
I .45-5.14
2.56-22.22
1.02-1.08
1.314.05
6.05-36.95
I .45-6.17
2.57-14.11
4.80-54.20
-
* Relative risk (RR) refers to the influence of the risk factor on the outcome variable; the RR is
adjusted for each of the variables in the analysis. RR = I for risk factor grade 0. Data are from the most
severely involved distal interphalangeal joint. See Subjects and Methods for explanations of the
grading systems. 95% CI = 95% confidence interval; Kellgren = Kellgren/Lawrence scale; Narrowing
=joint space narrowing.
t Excludes 5 subjects with both joint space narrowing and sclerosis.
joint abnormalities seen as isolated findings at the start
of the study (Figures 2A and B). These cross-sectional
results suggest that osteophytes and/or joint space
narrowing might be the earliest radiographic indicators
of hand OA. To test this hypothesis, Cox's proportional hazards regression models were used to deter-
mine the importance of isolated osteophytes and joint
space narrowing for predicting the future development
of OA (Table 1). Because the incidence data had
shown age to be an important predictor of OA development (Figure 3), age was included in each regression
model. Independent of the effect of age, joint space
KALLMAN ET AL
1328
Kellgren/Lawrence
Osteophytes
--1 I
I
0
D
5
to
t5
m
Tlme (yaanl
,
B
,
m
;I,,
0
6
,
,
1s
m
Time (ycrrs)
10
,
2s
10
Figure 4. Life-table analysis showing the progression of each radiographic feature of hand osteoarthritis for the most severely involved
distal interphalangeal joint. Young subjects ( 4 0 years) are represented by the long-dashed line, middle-aged subjects ( 4 W9 years)
by the short-dashed line, and old subjects (260 years) by the solid
line. For each radiographic feature, disease progressed significantly
faster in old subjects than in young or middle-aged subjects (P <
0.01).
narrowing, in the absence of any other abnormality,
significantly predicted the development of definite
small osteophytes (RR = 1.94, 95% CI 1.26-2.98).
However, isolated joint space narrowing did not predict the development of doubtful osteophytes (Kellgredlawrence grade 1). Conversely, isolated doubtful
osteophytes (KellgredLawrence grade 1) predicted
the development of joint space narrowing (RR = 2.06,
95% CI 1.19-3.56).
Joint space narrowing and osteophytes, either
definite or doubtful, were both powerful predictors of
the development of subchondral sclerosis. Finally,
subchondral sclerosis in combination with joint space
narrowing and osteophytes were strong predictors of
the development of lateral deformity and cortical
collapse (data not shown).
OA progression from any starting level. Eflect of
age. Life-table analysis was used to evaluate the
progression of at least 1 grade for each of the individual radiographic features of OA (Figure 4). In the DIP
joints, each radiographic feature of OA progressed
significantly faster (P < 0.01) in the old subjects than
in middle-aged or young subjects. OA progressed at
the slowest rate in the young subjects. The radiographic features of OA in middle-aged subjects usually
progressed at a rate intermediate between young and
old subjects, with the exception of joint space narrowing, which progressed at approximately the same rate
as it did in young subjects. For example, the median
time for 50% of the cohort to progress 1 Kellgred
Lawrence grade was 8.9 years for old subjects, 12.4
years for middle-aged subjects, and 15.8 years for
young subjects. Osteophytes and joint space narrowing progressed more slowly than did the changes
defined by the KellgredLawrence scale. Sclerosis
developed even more slowly, followed by lateral deformity and cortical collapse. Life-table analysis was
performed similarly on the PIP, CMCl, and TS joints.
Each feature progressed much more slowly in these
joints than in the DIP joints, but the pattern of OA
progressing the fastest in old subjects continued (data
not shown).
Proportional hazards models also showed age
to be an important predictor of OA progression from
any starting level. For instance, age increased the risk
of progression of both osteophytes and joint space
narrowing by 5% for each year of life (RR = 1.05,95%
CI 1.03-1.07). Thus, a 60-year-old subject had twice
the risk of osteophyte progression as a 40-year-old
subject had.
Efect of starting level. Cox's proportional hazards regression models were used to determine the
effect of the presence and severity of a feature at entry
on the progression of the feature. Because of the
known effect of age on progression (vide supra), age
was added to the regression equations as a continuous
variable. Two subjects who had severe disease (Kellgren/Lawrence grade 4), and were thus not at risk for
progression, were excluded from this analysis. For the
KellgrenlLawrence scale, the presence of moderate or
large osteophytes reduced the risk of osteophyte progression by 50% and 75%, respectively (RR = 0.46,
95% CI 0.26-0.82 and RR = 0.21, 95% CI 0.05-0.84,
respectively). Similarly, the presence of moderate
and/or severe joint space narrowing reduced the risk of
1329
COURSE OF HAND OA IN MALES
Table 2.
Cox’s proportional hazards regression models for the progression of existing osteoarthritis*
Outcome
variable
Risk factor
Osteophytes
(n = 27)
Age
Osteophytes
Narrowing
Grade
I
2
0
1
Sclerosis
2
3
0
1
Kellgren
(n = 90)
Age
Kellgren
Narrowing
1
2
3
0
1
2
Sclerosis
3
0
1
Narrowing
(n = 79)
Age
Narrowing
Osteoph ytes
I
2
0
Sclerosis
2
3
0
1
1
NO. of
subjects
at risk
NO. of subjects with
change
22
5
6
14
7
0
20
7
14
I
2
8
5
0
63
21
6
33
40
17
0
78
I2
54
14
2
23
31
16
0
61
9
1.04
I
0.46
0.20
60
38
8
31
1.01
I
0.32
I
1.31
3.62
4.30
19
57
17
4
I
66
13
10
5
10
4
I
36
10
RR
I .06
I
0.26
1
2.58
4.07
I
0.63
1
I .38
2.68
1
0.58
1
4.37
95% CI
0.96-1.15
0.03-1.98
0.53-12.58
0.50-32.86
0.12-3.26
I .02-1.07
0.27-0.90
0.05-0.90
-
0.79-2.40
1.28-5.61
0.25-1.35
0.97-1.05
0.124.84
-
0.57-3.02
1.24-10.56
0.45-40.77
1.53-12.49
* See Table 1 for explanations and definitions.
joint space narrowing progression by 50% and 70%,
respectively (RR = 0.50, 95% CI 0.32-0.77 and RR =
0.30, 95% CI 0.14-0.63). For both osteophytes and
joint space narrowing, more severe findings at the start
of the study decreased the risk of progression.
Risk factors for progression of existing OA.
Cox’s proportional hazards regression models were
also used to predict the progression of existing disease
(grade 1 or higher) based on a subject’s age and other
radiographic abnormalities (Table 2). For definite osteophytes, none of the independent covariates was
found to be a significant predictor of disease progression. There were only a small number of subjects,
however, who had definite osteophytes at entry into
the study and who were therefore at risk for progression (n = 27). These findings suggest the possibility of
Type I1 error.
Nonetheless, the relative risks for the association of joint space narrowing with progression of
osteophytosis are impressive: 2.58 and 4.07 for grade 1
and grade 2 narrowing, respectively. In addition, there
was a linear increase in the proportion of subjects
whose osteophytes progressed: 33%, 57%, and 71%
for grades 0, 1, and 2 of joint space narrowing,
respectively.
The KellgredLawrence scale confirmed the
previously described slowing of OA progression with
more severe disease and the increased progression of
OA with increasing age. The risk of joint space narrowing progression was significantly affected by other
abnormalities. The presence of moderate-sized osteophytes and subchondral sclerosis independently increased the risk of joint space narrowing progression.
Small osteophytes did not predict the progression of
joint space narrowing. Interestingly, age was not an
independent risk factor for the progression of joint
space narrowing, after an adjustment was made for
osteophytes and sclerosis in this model.
1330
KALLMAN ET AL
Early Stage
1
andlor
1
PartialJoint
1 H
Middle Stage
Small and moderate
osteophytes
Partialjoint narrowing
End Staae
1
SLOW
--
Large osteophytes
Joint space obliteration
Subchondral sclerosis
Lateral deformity
-HRng
Figure 5. Representation of the natural history of the radiographic features of osteoarthritis (OA),
based on our results. The middle and end stages correspond to grades 3 and 4 of the Kellgred
Lawrence scale. The early stage differs from the KellgredLawrence scale, in which doubtful
osteophytes and isolated joint space narrowing are not graded as evidence of OA.
DISCUSSION
In this longitudinal study, we used newly developed grading scales for the individual radiographic
features of osteoarthritis (21) to examine the natural
history of hand OA, and to test whether conventional
hypotheses about OA progression, based on prior
cross-sectional studies, were consistent with longitudinal findings. Compared with the global Kellgren/
Lawrence scale, these scales had the advantage of
being able to describe each radiographic abnormality
separately and in conjunction with other abnormalities. We found that the prevalence, incidence, and rate
of progression of each of the features of hand OA, as
measured by all scales, increased with age. We also
found previously undescribed relationships between
early stages of osteophytes and joint space narrowing,
each acting as risk factors for the development of OA.
The prevalence of hand OA (as judged by the
KellgredLawrence scale) in our study population was
consistent with the findings of other epidemiologic
surveys, which also demonstrated an increasing prevalence of hand OA across the age span (1-7). The
distribution of OA among the hand joints was similar
to that reported by Plato and Norris ( I ) , who studied a
larger cross-sectional sample of volunteers from the
Baltimore Longitudinal Study of Aging, as well as the
distribution found in the New Haven, CT (29) and
Tecumseh, MI (7) populations. In each report, DIP
joints had a higher prevalence of OA than did PIP
joints, with the second and fifth DIP joints being the
most frequently involved. The prevalence of OA in the
TS joints had previously been described only in case
reports (30). Thus, our results may be generalizable to
other male populations, given the comparability of
baseline prevalence data to these other populations.
However, we are unable to generalize these results
and their implications to women, since women were
not included in this analysis.
The incidence of hand OA in all joints increased
with age. In subjects older than 60, there was -1 new
case with doubtful osteophytes or joint space narrowing during each 10 person-years of followup. This
incidence rate may have been underestimated because
of the method we used to determine time at risk.
Subjects who previously had negative findings were
considered to be at risk until we observed a radiographic abnormality. Realistically, abnormalities
developed between the time radiographs were performed. If this hypothetical exact date of OA development was used, the number of years at risk would
have decreased, and incidence rates would have been
even higher.
We found that OA progressed very slowly. It
took approximately 10-20 years for any of the features
to progress 1 grade. It took even longer for subjects to
progress from the intermediate to the late stages of
OA, a possible “burnout” phenomenon. It was possible that this declining rate of progression may have
been a result of the scales rather than reflecting an
actual slowing of OA progression. If the Kellgred
Lawrence and joint space narrowing scales were more
widely spaced for more severe disease (i.e., grade 3
farther from grade 2 than grade 2 was from grade 1)
this would have led to the appearance of “burnout.”
However, there is evidence that supports the accuracy
of our observation of disease ‘‘burnout.” In a dog
model of knee OA, Sabiston et a1 (15) observed that
osteophytes, measured both by magnetic resonance
imaging and by plain radiography, initially grew very
quickly, and then their growth rate decreased. Slowing
of OA progression at higher levels has important
implications for clinical OA trials in which radiographic changes are used as outcome variables.
The Venn diagrams (Figures 2A and B) described interesting information about the prevalence of
isolated abnormalities and how these abnormalities
clustered. Because doubtful osteophytes (Kellgrenl
Lawrence grade 1) and joint space narrowing were the
most frequent findings, it was tempting to consider
them to be the initial findings of OA. Subchondral
1331
COURSE OF HAND OA IN MALES
sclerosis, because it was never seen without joint
space narrowing or without joint space narrowing and
osteophytes together, could be thought of as a later
finding. These hypotheses, based on cross-sectional
data, were supported by our longitudinal analysis,
which demonstrated that doubtful osteophytes, as a
lone finding, increased the risk of developing joint
space narrowing, and joint space narrowing increased
the risk of developing definite small osteophytes.
If we considered the simultaneous presence of
osteophytes and joint space narrowing to be convincing evidence of OA, then it is reasonable to conclude
that doubtful osteophytes and joint space narrowing,
occurring both alone and together, are the first radiographic signs of hand OA. Furthermore, osteophytes
and joint space narrowing, both alone and together,
were powerful risk factors for the development of
subchondral sclerosis, lending support to the hypothesis that subchondral trabecular remodeling was a
later change. Lateral deformity and cortical collapse
were powerfully predicted by the presence of osteophytes, joint space narrowing, and subchondral sclerosis, indicating that these were late or “end-stage’’
findings. The progression of moderate joint space
narrowing to joint obliteration was predicted by moderate osteophytes and subchondral sclerosis, and indicated that joint obliteration was a late finding. The
natural history of hand OA as implied by these data is
depicted in Figure 5 .
Radiography may be more sensitive t o some joint
abnormalities than others. Bone can be directly visuaiized, but cartilage can only be inspected indirectly, via
the width of the joint. Thus, any implications about the
pathophysiology of hand OA, based on the results of this
study, should be considered with caution. Nonetheless,
it is interesting to speculate about the pathophysiologic
implications of our findings.
Cartilage degeneration is accepted by most investigators as the first stage of OA. Our results corroborated this view. Isolated joint space narrowing
was a frequent finding, which predicted the development of osteophytes and sclerosis. It is interesting that
lone doubtful osteophytes were also an early radiographic indicator of hand OA. While it is possible that
osteophytes are a pathologic first phase of OA, they
might be reflective of cartilage damage that is not yet
identifiable as joint space narrowing on radiographs.
The other main hypothesis concerning the
pathophysiology of OA is that changes in subchondral
bone are the primary defect (31,32). Our results,
however, showed that sclerosis was never seen as a
sole finding and was not a predictor of the development of joint space narrowing. This may be explained
by a lack of sensitivity of plain radiographs to detect
early subchondral changes. A recent study was able to
demonstrate early subchondral bone changes in a
guinea pig model of OA, using microscopic computed
tomography (33).
The natural history of radiographic changes of
hand OA in men, as described by these results, has
important implications for epidemiologic studies of
hand OA. Previous studies have traditionally considered subjects to have OA if they had at least grade 2
changes (definite osteophytes) by the Kellgrenl
Lawrence scale, and have not considered doubtful
osteophytes and isolated joint space narrowing as
being diagnostic of OA. Our longitudinal data showed
that doubtful osteophytes and joint space narrowing,
occurring together or separately, were important predictors of OA development and progression. Together,
these results suggest a model for the natural history of
radiographic changes in OA of the hand. Future longitudinal, population-based epidemiologic studies,
such as the Third Health and Nutrition Examination
Survey, currently in progress, may provide data that
validate this model.
ACKNOWLEDGMENTS
We thank Harry Carr, James Wood, and Kim Roadarmel for performing the hand radiography, Kim Roadarmel
for technical assistance. the subjects of the Baltimore Longitudinal Study of Aging for their devotion to research on
aging, and Karen Taylor for secretarial expertise.
REFERENCES
1. Plato CC, Norris AH: Osteoarthritis of the hand: agespecific joint-digit prevalence rates. Am J Epidemiol
109: 169-180, 1979
2. Lawrence LS, Bremmer JM, Bier F: Osteo-qrthrosis:
prevalence in the population and relationship :between
symptoms and x-ray changes. Ann Rheum Dis 25: 1-23,
1966
3. Roberts J, Burch TA: Prevalence of osteoarthritis in
adults. Vital Health Stat [ l l ] , No. 15, 1966
4. Caird FI, Webb J, Lee P: Osteoarthrosis of the hands in
the elderly. Age Ageing 2: 150-156, 1973
5 . Mikkelsen WM, Duff IF, Dodge HJ: Age-sex specific
prevalence of radiographic abnormalities of the joints of
the hands, wrists and cervical spine of adult residents of
the Tecumseh, Michigan, community health study area,
1962-1965. J Chronic Dis 23:151-159, 1970
1332
6. Soila P, Pyykonen L: Tables of the incidence of osteochondrosis in joints. Scand J Rheumatol6:151-160, 1960
7. Butler WJ, Hawthorne VM, Mikkelsen WM, Carman
WJ, Bouthillier DL, Lamphiear DE, Kazi IU: Prevalence of radiologically defined osteoarthritis in the finger
and wrist joints of adult residents of Tecumseh, Michigan, 1962-65. J Clin Epidemiol 41:467473, 1988
8. Plato CC, Norris AH: Osteoarthritis of the hand: longitudinal studies. Am J Epidemiol 110:740-746, 1979
9. Busby J, Tobin JT, Ettinger W, Plato CC: Progression of
osteoarthritis: significance of starting level, age, and
length of follow-up (abstract). Gerontologist 26 (suppl):
141A, 1986
10. Brandt KD, Fife RS: Ageing in relationship to the
pathogenesis of osteoarthritis. Clin Rheum Dis 12: 117130, 1986
11. Hall FM, Wyshak G: Thickness of articular cartilage in
the normal knee. J Bone Joint Surg 62A:408413, 1980
12. Sartoris DJ, Resnick D: Radiological changes with ageing in relation to bone disease and arthritis. Clin Rheum
Dis 12:181-226, 1986
13. Kellgren JH, Lawrence JS: Radiological assessment of
osteoarthrosis. Ann Rheum Dis 16:494-501, 1957
14. Kellgren JH: The Epidemiology of Chronic Rheumatism. Second edition. Philadelphia, FA Davis, 1963
15. Sabiston CP, Adams ME, Li DKB: Magnetic resonance
imaging of osteoarthritis: correlation with gross pathology using an experimental model. J Orthop Res 5:164172, 1987
16. Moskowitz RW, Goldberg VM: Osteophyte evolution:
studies in an experimental partial meniscectomy model.
J Rheumatol [Suppl] 14:116-118, 1987
17. Ahlback S: Osteoarthritis of the knee: a radiographic
investigation. Acta Radio1 [Suppl] (Stockh) 277:7-7 I ,
1968
18. Hernborg J, Nilsson BE: The relationship between
osteophytes in the knee joint, osteoarthritis and aging.
Acta Orthop Scand 44:69-74, 1973
19. Danielsson L, Hernborg J: Clinical and roentgenologic
study of knee joints with osteophytes. Clin Orthop
69~302-312, 1970
KALLMAN ET AL
20. Danielsson LG: Incidence and prognosis of coxarthrosis. Acta Orthop Scand [Suppl] 669-87, 1964
21. Kallman DA, Wigley FM, Scott WW Jr, Hochberg MC,
Tobin JD: New radiographic grading scales for osteoarthritis of the hand: reliability for determining prevalence
and progression. Arthritis Rheum 32: 1584-1591, 1989
22. Shock NW, Gruelich RC, Andres RA, Arenberg D,
Costa PT, Lakatta EG, Tobin JD: Normal Human
Aging: The Baltimore Longitudinal Study of Aging.
Washington, DC, US Government Printing Office, 1984
23. Bartko JJ: The intraclass correlation coefficient as a
measure of reliability. Psycho1 Rep 19: 1-1 1, 1966
24. SAS Institute Inc: SAS User’s Guide: Statistics. Version 5 . Cary, NC, SAS Institute, 1985
25. Hopkins A: Survival analysis with covariates: Cox models, BMDP Statistical Software Manual. Edited by WJ
Dixon, MB Brown, L Engelman, JW Frane, MA Hill, RI
Jennrich, JD Toporek. Berkeley, University of California Press, 1985
26. Lawless JF: Statistical Models and Methods for Lifetime Data. New York, John Wiley & Sons, 1982
27. Cox DR: Partial likelihood. Biometrika 62:269-276, 1975
28. Breslow N: Covariance analysis of censored survival
data. Biometrics 30:89-99, 1974
29. Acheson RM, Chan YK, Clemett AR: New Haven
survey ofjoint diseases. XII. Distribution and symptoms
of osteoarthrosis in the hands with reference to handedness. Ann Rheum Dis 29:275-285, 1970
30. Patterson AC: Osteoarthritis of the trapezioscaphoid
joint. Arthritis Rheum 18:375-379, 1975
31. Radin EL, Paul IL, Tolkoff MJ: Subchondral bone
changes in patients with early degenerative joint disease.
Arthritis Rheum 13:400-405, 1970
32. Radin EL, Rose RM: Role of subchondral bone in the
initiation and progression of cartilage damage. Clin
Orthop 213:34-40, 1986
33. Layton MW, Goldstein SA, Goulet RW, Feldkamp LA,
Kubinski DJ, Bole GG: Examination of subchondral
bone architecture in experimental osteoarthritis by microscopic computed axial tomography. Arthritis Rheum
31:1400-1405, 1988
Документ
Категория
Без категории
Просмотров
4
Размер файла
838 Кб
Теги
course, population, hands, malen, osteoarthritis, longitudinal
1/--страниц
Пожаловаться на содержимое документа