Population
The study population was a stratified two-stage cluster sample of 8000
persons drawn from the population register to represent Finnish adults thirty
years of age or
older19. In the
first stage, forty representative areas were selected. In the second stage, a
systematic sample of inhabitants was drawn from each area. The sample
consisted of 8000 persons (3637 men and 4363 women) from sixty-nine
municipalities.
Baseline Examination
The baseline examinations were carried out from 1978 through 1980 by the
Mobile Clinic of the Social Insurance Institution. A total of 7217 subjects
(90% of the sample) participated in a screening phase. The distribution of
sex, age, and level of education among the participants corresponded closely
to that of the whole Finnish
population19. The
examinations included administration of questionnaires, interviews about
symptoms, and various measurements and tests.
The individuals who screened positive—i.e., those with a history,
symptoms, or findings suggestive of musculoskeletal disease—as well as a
20% subsample of those attending the screening phase were invited to
participate in the diagnostic (clinical) phase of the study. In the clinical
phase, a posteroanterior radiograph of the hand was made of 3595 subjects,
2503 of whom had been invited to participate because they had screened
positive for musculoskeletal disease and 1092 of whom were in the random 20%
subsample. Of the latter group, 627 screened positive and 465 screened
negative for musculoskeletal disease. The 3595 subjects constituted the study
population that was followed to determine disability and mortality.
The methods for studying musculoskeletal diseases are described in detail
in the Appendix and in previous
articles18,20-23.
A specific interview conducted by specially trained nurses included questions
about musculoskeletal symptoms and disorders and their consequences. A
standardized joint function test was carried out to screen for musculoskeletal
impairment. Overall disability was defined as at least marked difficulty in
moving about the house, getting in and out of bed, dressing and undressing,
walking 400 m, carrying a shopping bag weighing 5 kg, climbing a flight of
stairs, grocery shopping, clipping toenails, reading a newspaper, traveling on
public transportation, or performing strenuous housework such as
cleaning22.
Physical examinations of 3434 subjects were carried out by specially
trained physicians according to a standardized written protocol to grade
deformity, mobility, and tenderness of all limb joints, including the two
thumb carpometacarpal joints separately. Chronic diseases were diagnosed on
the basis of the medical history, symptoms, and clinical
findings19-23.
Radiographic Methods
The posteroanterior radiographs of the hand were evaluated by a radiologist
without any information on the subject's clinical
findings24.
Standard
criteria25,26
were used to assess joint space narrowing and osteophytes. Each joint of both
hands was graded according to the method of Kellgren et
al.25, in which
grade 0 indicates no osteoarthritis; grade 1, minimal osteophytes and possible
cyst formation (doubtful osteoarthritis); grade 2, definite osteophytes and
possible cysts (minimal osteoarthritis); grade 3, moderate osteophytes,
narrowing of the joint space, subchondral sclerosis, and deformity of the bone
ends (moderate osteoarthritis); and grade 4, large osteophytes, severe
sclerosis, and narrowing of the joint space (severe osteoarthritis). In the
present study, grade 2, 3, or 4 was considered to denote thumb carpometacarpal
osteoarthritis, whereas grade 3 or 4 was designated as advanced thumb
carpometacarpal osteoarthritis.
Definition of Risk Factors
A basic questionnaire was used to elicit information about present and
previous occupations involving lifting or carrying heavy objects; stooped,
twisted, or otherwise awkward work postures; vibration of the whole body or
use of vibrating equipment; continuously repeated series of movements; and
work paced by a machine. Only exposures related to the present or to the last
occupation were taken into account. These exposures were recorded as
dichotomies (no = 0 and yes = 1), and the total number was designated as
"the sum index of physical stress at
work."21
Standing height and weight were measured at the screening examination, and
the body mass index (weight divided by the square of the height
[kg/m2]) was used as a measure of relative weight. Smoking history
was determined during a standard interview and was categorized as never
smoked; quit smoking; current smoker of cigars, pipe, or fewer than twenty
cigarettes a day; or current smoker of twenty cigarettes or more a day. The
basic questionnaire inquired about average weekly consumption of beer, wine,
and strong alcoholic beverages during the preceding month. The overall alcohol
consumption was then calculated and expressed in grams of ethanol per week.
The level of education was considered in three categories based on years of
education.
Follow-up
The morbidity and mortality of the subjects were followed systematically
after the baseline examination. To identify new cases of work disability, data
on new national disability pensions were collected from the Social Insurance
Institution's register. Information on mortality was obtained from the Central
Statistical Office of Finland. Both the work disability and the mortality are
reported up to 1995—i.e., over a period of up to seventeen years.
Statistical Methods
Statistical Analysis System (SAS) software (SAS Institute, Cary, North
Carolina) was used for the statistical analysis. Adjusted prevalence rates of
thumb carpometacarpal osteoarthritis were estimated with use of a general
linear model27.
Differences in prevalence between subgroups or between levels of suspected
risk determinants and their first-degree interaction terms were analyzed with
a logistic regression
model28. Relative
risks were estimated as adjusted odds ratios. The Cox lifetable regression
model was used to study the ability of thumb carpometacarpal osteoarthritis to
predict work disability and
mortality29.
Potential confounding and effect-modifying factors were also entered into the
models. Significance of interaction was tested by entering interaction terms
into the models. The 95% confidence intervals of the relative risk estimates
and the likelihood ratio statistics (chi-square values expressed as p values)
were based on the models.
Prevalence of Thumb Carpometacarpal Osteoarthritis
Figure 1 shows the
prevalences of thumb carpometacarpal osteoarthritis of Kellgren grade 2, 3, or
4 and of Kellgren grade 3 or 4 by sex and age. The prevalence of symmetrical
thumb carpometacarpal osteoarthritis was 6.0% in women and 2.6% in men.
Determinants of Thumb Carpometacarpal Osteoarthritis
In addition to age and female gender, the body mass index was a strong
determinant of thumb carpometacarpal osteoarthritis (see Appendix). After
adjusting for age, gender, and other alleged risk determinants, we found the
body mass index to be directly proportional to the prevalence of thumb
carpometacarpal osteoarthritis in both sexes. The adjusted odds ratio was 1.29
(95% confidence interval, 1.15 to 1.43) per 5-kg/m2 increment in
body mass index. Thus, the subjects with a body mass index 35 kg/m2
had more than a twofold risk of having thumb carpometacarpal osteoarthritis
compared with those with a normal relative weight (body mass index of 20.0 to
24.9 kg/m2). However, the body mass index showed a weaker
association with advanced thumb carpometacarpal osteoarthritis (Kellgren grade
3 or 4), with an adjusted odds ratio of 1.14 (95% confidence interval, 0.92 to
1.37) per 5-kg/m2 increment in body mass index.
Neither the level of education nor the physical workload history showed a
clear-cut association with the prevalence of thumb carpometacarpal
osteoarthritis (see Appendix).
There was a significant difference between men and women with regard to the
association between smoking and thumb carpometacarpal osteoarthritis (p value
for interaction, 0.004). The association was positive for women whereas, among
men, both past and current smoking was negatively associated with the
prevalence of thumb carpometacarpal osteoarthritis (see Appendix).
The prevalence of thumb carpometacarpal osteoarthritis of Kellgren grade 2,
3, or 4 was not significantly associated with alcohol consumption. Advanced
(grade-3 or 4) thumb carpometacarpal osteoarthritis, however, showed a
significant association with intakes of 250 g of ethanol per week by men (odds
ratio, 1.32; 95% confidence interval, 1.03 to 1.69). Otherwise, the patterns
of risk determinants did not differ between osteoarthritis of Kellgren grade
2, 3, or 4 and osteoarthritis of Kellgren grade 3 or 4.
Concomitant Diseases
Thumb carpometacarpal osteoarthritis had a strong association with other
types of hand osteoarthritis. After adjustment for age and gender, there was
an association between osteoarthritis (Kellgren grade 2, 3, or 4) in any
finger joint and osteoarthritis in the thumb carpometacarpal, with an odds
ratio of 8.35 (95% confidence interval, 5.30 to 13.2). A weaker association,
with an odds ratio of 3.66 (95% confidence interval, 2.85 to 4.70), was found
between thumb carpometacarpal osteoarthritis and symmetrical osteoarthritis in
the distal interphalangeal joints. However, thumb carpometacarpal
osteoarthritis showed no significant association with other clinically
diagnosed chronic diseases such as knee osteoarthritis, hip osteoarthritis,
neck pain, low back pain, diabetes, or cardiovascular disease.
Physical Status
There was a strong association between thumb carpometacarpal osteoarthritis
and the physical status of the carpometacarpal joint of the ipsilateral thumb,
including restriction of movement, pain with movement, swelling, and
tenderness. Subjects with any of these findings had a threefold risk of having
radiographic signs of osteoarthritis in the right hand (odds ratio, 3.29; 95%
confidence interval, 2.03 to 5.33) and a twofold risk in the left hand (odds
ratio, 2.16; 95% confidence interval, 1.34 to 3.51).
Disability
At baseline, thumb carpometacarpal osteoarthritis was not associated with
overall disability. After adjustment for gender and age, the odds ratio for
the inability to perform ordinary daily activities in the presence of thumb
carpometacarpal osteoarthritis was 0.80 (95% confidence interval, 0.63 to
1.01). Thumb carpometacarpal osteoarthritis also did not predict the incidence
of work disability: after adjustment for age and gender, the relative risks of
work disability during the follow-up period were 0.91 (95% confidence
interval, 0.61 to 1.38) and 1.47 (95% confidence interval, 0.65 to 3.31) for
patients with thumb carpometacarpal osteoarthritis of Kellgren grade 2, 3, or
4 and those with grades 3 or 4, respectively.
Mortality
After adjustment for other potential confounding factors, no significant
association was found between thumb carpometacarpal osteoarthritis of Kellgren
grade 2, 3, or 4 and mortality with any cause or mortality with a
cardiovascular cause. However, men who had advanced thumb carpometacarpal
osteoarthritis of Kellgren grade 3 or 4 had a slightly increased risk of
mortality with any cause; the adjusted relative risk was 1.32 (95% confidence
interval, 1.03 to 1.69).
Our results confirm that thumb carpometacarpal osteoarthritis is a common
finding in the general Finnish population and is clearly more common in women
than in men. The prevalence of thumb carpometacarpal osteoarthritis was
closely associated with obesity, a finding that is in agreement with those of
other reports on osteoarthritis in hand
joints12,14,18,30.
The level of education and the physical work-load history showed no
significant association with thumb carpometacarpal osteoarthritis. In
addition, we found the effect of thumb carpometacarpal osteoarthritis on
disability to be modest compared with the effect reported in clinical series
of patients or studies of elderly
subjects3,5.
One of the main findings in the present study was a positive association
between thumb carpometacarpal osteoarthritis and body mass index. A similar
association has been reported in previous
studies12,31.
Basically, two different mechanisms have been proposed to explain these
findings32.
According to the mechanical theory, obesity causes osteoarthritis by
increasing the load and impact on the joint, which, in turn, promotes
degeneration of the cartilage. Alternatively, hormones or other biological
factors related to obesity may predispose to or cause
osteoarthritis32.
Although some studies have suggested that insulin-like growth factor and sex
hormones might exert a favorable effect on the metabolism of articular
cartilage33-36,
contrasting results have also been
reported37,38.
Some studies have even suggested that high levels of serum lipids can
accelerate the degradation of articular
cartilage39-41.
Obesity has been found to be associated with osteoarthritis in joints that are
not weight-bearing, such as the distal interphalangeal
joints18 and the
carpometacarpal
joints12. One
previous study also demonstrated increases in the risks of tibiofemoral,
patellofemoral, and carpometacarpal osteoarthritis with increasing body
weight12.
Therefore, obesity appears to affect the joint cartilage both mechanically and
systemically.
The associations between smoking and thumb carpometacarpal osteoarthritis
differed significantly between men and women. In men, smoking seemed to be
associated with a reduced risk of thumb carpometacarpal osteoarthritis,
whereas the opposite was seen in women. This finding is, however, far from
being conclusive because of the small number of female smokers. One previous
study revealed no obvious protective effect of smoking on hand and knee
osteoarthritis in
women42, whereas
another study showed lower rates of knee osteoarthritis in male and female
smokers compared with the rates in
nonsmokers43. Our
results are consistent with those of our earlier study on finger
osteoarthritis and its association with
smoking18. It is
not known whether smoking affects cartilage or bone, or both. Smoking may not
be the only explanatory variable; part of the observed favorable effect of
smoking might be explained by other lifestyle factors.
Mechanical stress and biomechanical factors are considered to be important
risk factors for thumb carpometacarpal osteoarthritis, although there is no
direct evidence that they
are11,13,14.
Segal et al.44
found less osteoarthritis in paralytic hand joints than in nonparalytic hand
joints in their elderly patients, which supports the role of biomechanical
factors in the development of osteoarthritis. We previously reported that
manual labor is a risk factor for finger osteoarthritis in women but not in
men18. In the
present study, however, the risk of thumb carpometacarpal osteoarthritis was
not associated with either the physical work stress index or the level of
education of either sex. Similarly, Jones et
al.16 found no
association between thumb carpometacarpal osteoarthritis and occupation or
past physical activity in their clinical study. Thus, mechanical stress
apparently contributes to the development of osteoarthritis in many joints but
not necessarily in the thumb carpometacarpal joint.
An important finding of our study was that even if thumb carpometacarpal
osteoarthritis is common, it rarely causes disability; this was observed in
both cross-sectional and longitudinal studies. These results disagree with
those reported by
others3,5,
perhaps as a result of differences in study populations. We studied subjects
from a nationally representative population sample of individuals thirty years
of age or older, whereas in the other studies the subjects were selected from
clinical sources or from elderly age groups. Work disability due to thumb
carpometacarpal osteoarthritis is likely in jobs with physically high demands
on the hands, but such an association cannot be detected in study populations
involving only subjects at high risk.
A previous report suggested that thumb carpometacarpal osteoarthritis is
associated with an increased risk of osteoarthritis in the metacarpophalangeal
joints and, to a lesser extent, in the distal interphalangeal and proximal
interphalangeal
joints45. In our
study, thumb carpometacarpal osteoarthritis was usually symmetrical and had a
strong association with other types of hand osteoarthritis, suggesting that
genetic factors are involved in the development of osteoarthritis of the hand.
While associations between osteoarthritis in the thumb carpometacarpal and
osteoarthritis in weight-bearing joints have been found
previously7,8,46,
we did not observe such an association.
In conclusion, obesity is a strong determinant of thumb carpometacarpal
osteoarthritis in both sexes. The effect of thumb carpometacarpal
osteoarthritis on disability and mortality in the general population is
modest. Because of the rarity of ensuing disability, thumb carpometacarpal
osteoarthritis is likely to be underdiagnosed in clinical practice.
A description of the screening procedure for musculoskeletal disorders used
in this study and a table depicting all of the correlations studied are
available with the electronic versions of this article, on our web site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).