The Disabilities of the Arm, Shoulder and Hand (DASH)
questionnaire1 is
the most widely used upper extremity-specific outcome
instrument2. The
wide variations observed in DASH scores for patients with specific conditions
are difficult to explain on the basis of variations in objective pathology
alone3. Furthermore,
pain in the absence of objective dysfunction can be as disabling as conditions
that cause obvious
impairment3.
Psychological factors such as pain anxiety (a mental state of perceived danger
that is provoked by the experience or anticipation of
pain)4 and pain
catastrophizing5-7
(a maladaptive coping mechanism through which patients cognitively address
their pain, thereby intensifying the fearful aspects of the experience of
pain) have been associated with nonspecific or idiopathic arm
pain8.
Neuroticism, sometimes called trait negative affect or emotionality, is the
stable tendency toward negative mood, negative self-concept, distress, and
dissatisfaction9.
Neuroticism is a personality trait that has been linked with self-reported
physical symptoms in a number of
populations10. The
trait has been correlated with the reporting of several common health
complaints (for example, stomach aches and joint
pain)9,11.
Other studies have demonstrated that neuroticism was linked to the reporting
of symptoms with no physiological basis in patients with
colds12. Among
patients with major depression, neuroticism helped to explain discrepancies
between self-reported and observer-reported depression symptom
severity13.
In the present study, we tested the hypothesis that, accounting for the
influence of depression and anxiety, the personality trait of neuroticism
correlates with DASH scores.
Patients who presented for the treatment of carpal tunnel syndrome,
unilateral de Quervain tendinitis, a single trigger finger, or unilateral
lateral elbow pain, and patients who were recovering from a fracture of the
distal part of the radius that had been treated nonoperatively six weeks
previously were identified and were invited to participate in the study by
their treating physician in the office setting. Patients with multiple
diagnoses and those in whom the diagnosis was questionable were excluded. The
study was performed with use of a protocol that had been approved by our Human
Research Committee. Informed consent was obtained from each patient.
Patients
Two-hundred and thirty-five patients were enrolled, including forty-five
patients who had carpal tunnel syndrome, forty-four who had de Quervain
tendinitis, forty-eight who had lateral elbow pain, seventy-one who had a
single trigger finger, and twenty-seven who were recovering from a distal
radial fracture that had been treated nonoperatively. The demographic
characteristics of each cohort were comparable overall, with the exception of
some variability in terms of the average age
(Table I). Of the 235 patients,
fifty-nine (25%) were male and 176 (75%) were female. Gender was comparable
among the cohorts (chi square = 4.37, degrees of freedom = 4, p = 0.36). Age
differed significantly between cohorts (F = 13.87, degrees of freedom = 4, p
< 0.01), with patients who had a diagnosis of trigger finger or a distal
radial fracture being significantly older on the average. The average age (and
standard deviation) of all patients was 52.2 ± 17.8 years (range,
twenty-one to eighty-six years). Work status (chi square = 25.82, degrees of
freedom = 20, p = 0.17), marital status (chi square = 16.45, degrees of
freedom = 12, p = 0.17), and educational level (chi square = 11.91, degrees of
freedom = 8, p = 0.16) were comparable among cohorts.
Measures
The Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire was
used to measure self-reported upper extremity-specific health
status1. The DASH
questionnaire was developed by the American Academy of Orthopaedic Surgeons in
conjunction with the Council of Musculoskeletal Specialty Societies and the
Institute for Work and
Health1. Scores
range from 0 to 100, and the average DASH score (and standard deviation) for
the general population was reported to be 10.10 ± 14.68 by Hunsaker et
al.2.
Neuroticism was assessed with use of a validated instrument, the Eysenck
Personality Questionnaire-Revised
(EPQ-R)14. The
EPQ-R is one of the most widely used personality inventories, in part because
of the uniformity of responses between
genders15.
Neuroticism scores range from 0 to 24.
Depressive symptoms were assessed with the Center for Epidemiologic
Studies-Depression (CES-D) scale. The CES-D is a valid and reliable measure of
depressive
symptoms16, and the
widely used predictive cutoff score of 16 can be used to estimate the
prevalence of Major Depressive
Disorder17. CES-D
scores range from 0 to 60, with an average score of 9.1 ± 8.6 for the
general
population17.
The Pain Anxiety Symptoms Scale (PASS) was used to assess anxiety about
pain4. The PASS is a
valid and reliable measure of four aspects of pain anxiety: (1) cognitive
anxiety (the inability to concentrate and the intrusion of unwanted thoughts
when in pain), (2) fear of pain (thoughts of dread and terror when in pain),
(3) escape and avoidance (behavior that minimizes the duration and amount of
pain), and (4) physiological anxiety (physical symptoms such as sweating and
shaking)4. The PASS
score ranges from 0 to 200 overall and from 0 to 50 on each of the
subscales.
Statistical Analysis
Continuous variables were assessed for normality with the
Kolmogorov-Smirnov test and were found to conform to a Gaussian-shaped
distribution18.
Therefore, the F test based on analysis of variance was used to compare age
and DASH scores between the five different cohorts and the chi-square test was
used to evaluate differences regarding categorical variables, including
gender, education, marital status, and professional work status.
The Pearson product-moment correlation coefficient (r) was calculated to
assess the relationship between the DASH score and the EPQ-R, PASS, and CES-D
scores. Analysis of covariance and multiple linear regression were applied to
determine independent predictors of the DASH score and to account for
potential confounding or redundancy among predictor variables. Once it was
clear that depression was the strongest influence on scores, a common-slope
model was used to derive equations for predicting the DASH score from the
CES-D score for each of the five diagnoses and 95% confidence intervals
(representing the normal range) were constructed around the predicted values
for the DASH score at each 10-point interval of the CES-D
score19.
The SPSS statistical package (version 13.0; SPSS, Chicago, Illinois) was
used for analysis and modeling of the data. Two-tailed p values of <0.05
were considered significant. A power analysis indicated that a minimum sample
size of twenty-five patients in each of the five study groups would provide
80% probability to detect a mean difference of 10 points in the DASH score
with use of analysis of variance and to estimate the true linear correlation
between the DASH score and psychological variables with a 95% confidence
interval around an assumed linear correlation of r = 0.40 (nQuery Advisor,
version 5.0; Statistical Solutions, Cork, Ireland).
The average DASH score (and standard deviation) was 29.0 ± 20.5 for
patients with carpal tunnel syndrome, 32.7 ± 17.9 for those with de
Quervain tendinitis, 33.3 ± 21.5 for those with lateral elbow pain,
22.0 ± 14.3 for those with trigger finger, and 34.8 ± 15.0 for
those with a distal radial fracture. The average DASH score for all patients
was 29.3 ± 18.6 (range, 1 to 92). The average DASH score was
significantly lower for patients with a trigger finger than for those with
other diagnoses (F = 4.45, degrees of freedom = 4, p < 0.01).
No significant correlation was observed between neuroticism and the DASH
score in any cohort (r = 0.24 and p = 0.10 for all)
(Table II). Significant
moderate correlation was observed between the DASH score and depression for
all diagnoses (p < 0.01 for all). The relationship between the DASH score
and pain anxiety varied for each cohort
(Table II).
In patients with carpal tunnel syndrome, significant correlations were
observed between the DASH score and depression (r = 0.50, p < 0.01),
pain-escaping behavior (r = 0.41, p < 0.01), cognitive anxiety (r = 0.37, p
< 0.05), physical symptoms when in pain (r = 0.39, p < 0.01), and the
total PASS score (r = 0.40, p < 0.01).
In patients with de Quervain tenosynovitis, significant correlations were
observed between the DASH score and depression (r = 0.49, p < 0.01), fear
of pain (r = 0.45, p < 0.01), somatic symptoms (r = 0.53, p < 0.01),
cognitive anxiety (r = 0.36, p < 0.05), pain-escaping behavior (r = 0.32, p
< 0.01), and the total PASS score (r = 0.46, p < 0.01).
In patients with lateral elbow pain, significant correlations were observed
between the DASH score and depression (r = 0.38, p < 0.01), pain-escaping
behavior (r = 0.55, p < 0.01), fear of pain (r = 0.32, p < 0.05),
cognitive anxiety (r = 0.28, p < 0.05), and the total PASS score (r = 0.42,
p < 0.01). In patients with trigger finger, significant correlations were
observed between the DASH score and depression (r = 0.47, p < 0.01),
cognitive anxiety (r = 0.38, p < 0.01), and, although the correlation was
limited, with the total PASS score (r = 0.24, p < 0.05).
In patients with a distal radial fracture, the only significant correlation
was between the DASH score and depression (r = 0.52, p < 0.01).
When all of the patients were analyzed together, the DASH score correlated
mildly to moderately with depression (r = 0.44, p < 0.01), pain-escaping
behavior (r = 0.30, p < 0.01), fear of pain (r = 0.22, p < 0.05),
cognitive anxiety (r = 0.27, p < 0.05), physical symptoms when in pain (r =
0.29, p < 0.01), and the total PASS score (r = 0.31, p < 0.01).
Analysis of covariance revealed a highly significant effect of depression
(as measured with the CES-D score) on the DASH score for all diagnoses, with
higher depression scores being predictive of higher (worse) DASH scores and
lower depression scores being predictive of lower (better) DASH scores. The
interaction between the covariate CES-D score and the diagnosis group was not
significant (F = 0.68, p = 0.61), indicating that the slope of the
relationship between the CES-D and DASH scores was equal for all five cohorts
(Fig. 1). Hence, further
modeling to derive predictive DASH scores on the basis of CES-D scores for
each cohort was performed with use of a regression model with a common slope
(Table III). Of interest,
although male and female patients showed no significant differences in
depression scores (p = 0.74, Student t test), their DASH scores differed
significantly (p = 0.003, Student t test).
Multivariate analysis indicated that both depression as measured with the
CES-D (F = 62.68, p < 0.0001) and gender (F = 11.36, p < 0.001) were
independent predictors of the DASH score, with females having higher DASH
scores than males for any given CES-D score within each of the five cohorts.
Education, marital status, and work status had no significant influence on the
DASH score (p > 0.25 for all).
The present study demonstrated that neuroticism does not correlate with the
DASH score but depression and pain anxiety do. Depression in particular was
found to correlate moderately with the DASH score for patients with a variety
of common arm problems. The five conditions that we studied are discrete,
easily defined problems with a relatively limited variation in objectively
measurable pathology or dysfunction. In this setting, our opinion is that the
large standard deviations of the DASH score (between 14.3 and 21.5) are not
adequately explained by objective, physical factors. Our data suggest that
this variability arises from psychosocial rather than physical factors. These
findings may be common to all health-status measures.
Evidence is mounting that self-assessed disability is related as much or
more to illness behavior than to
pathophysiology3,8.
In particular, adequate coping mechanisms are important in both the experience
of pain and the perception of
disability3.
Patients who suffer from depression or pain anxiety may not be as capable of
adapting to and managing painful upper extremity problems. They may see
themselves as more disabled than would be expected on the basis of physical,
objective findings.
"Damage" (e.g., a distal radial fracture) or
"malfunction" (e.g., lateral elbow pain, trigger finger, de
Quervain tendinitis, or carpal tunnel syndrome) may only be responsible for a
limited part of a patient's perceived disability. Psychosocial factors may be
responsible for a substantial part.
Our profession would not be well served or well represented if we tended to
operate or intervene largely because a person is vulnerable and is having
difficulty coping with his or her illness. We need to improve our ability to
identify objective, physical problems that merit operative intervention in
order to improve health status. In other words, we need to be cautious and
skeptical with regard to diagnoses and treatments that are not supported by
objectively verifiable data, and we need to try to restrict our practices to
treatments of proved efficacy for clearly defined diagnoses that are
verifiable with reproducible objective tests (often referred to as the
practice of evidence-based medicine). There is a need to identify and account
for subjective, psychological factors that keep patients from coping with
relatively harmless, common, and usually self-limiting problems such as
lateral elbow pain and de Quervain tenosynovitis. ?